Skip to main content

Full text of "The magnetite deposits of the eastern Mesabi Range, Minnesota"

See other formats



2ty* IntwrHttg of iitttttp00ta 

William H. Emmons, Director 






The University of Minnesota 

William H. Emmons, Director 






The University of Minnesota 




Chapter I. Introduction 1-4 

Location '. 

History 2 

Present exploration 3 

Summary of origin of the magnetite 3 

Chapter II. General geology of the East Mesabi range 5-11 

The formations 5 

The Lower-Middle Huronian series 5 

The slate and schist formation 5 

The Giants Range granite 5 


The Upper Huronian series (Animikie group) 7 

The Pokegama quartzite 

The Biwabik iron-bearing formation 7 

The Virginia slate 7 

The sills and dikes 8 

Keweenawan series 8 

The Duluth gabbro 8 

Contact metamorphism 8 

Regional metamorphism 10 

Structure 10-11 

Glaciation 11 

Weathering 11 

Chapter III. The material of the iron-bearing formation 11-15* 

Minerals of the iron-bearing formation 12 

Rocks of the iron-bearing formation 14 

Chapter IV. Subdivisions of the iron-bearing formation 16 

Lower cherty beds 18 

Lower slaty beds 19 

Upper cherty beds 19 

Upper slaty beds 22 

Correlation of sediments of east and west Mesabi 25 

Chapter V. The hematite ore deposits 27-28 

Chapter VI. The magnetite deposits — general 29-34 

The minerals of the deposits 29 

Chemical composition of the magnetite rock 29 

Textures 30 

Stratigraphy of the magnetite 31 

Distribution of the magnetite deposits 32 



Magnetic mapping 33 

Method of estimating reserves 34 

Chapter VII. The magnetite deposits — economic considerations. . 35-39 

Commercial situation 35 

Per cent of iron and volume of material 35 

Topographic situation 37 

Structures 37 

Magnetism 37 

Size of grain 3S 

Hardness 38 

Milling 38 

Grade of product 38 

Transportation 38 

Cost of exploration 39 

Scale of operations 39 

Suggestions for conservation 39 

Chapter VIII. The origin of the magnetite deposits 40-47 

Introduction 40 

Agent of transportation 40 

Source of the iron 40 

Deposition 40 

Primary modification of the deposit 42 

Metamorphism 45 

Lack of secondary enrichment 46 

Resume of the history 46 

Chapter IX. Detail of the districts 48-56 

The Mesaba area 48 

The Spring mine area 48 

The Ridge area 51 

The Jericho area 51 

The Iron Lake area 52 

The Sulphur area 53 

The Birch Lake area 54 


Plate I. General geologic map of the eastern Mesabi dis- 
trict in pocket 

II. Columnar section of the iron formation in the east 

Mesabi 4 

III. A. The basal conglomerate of the Biwabik iron for- 

mation, resting on the granite 

B. The lower cherty beds showing a structure like 

piles of inverted bowls 

C. Granules of quartz, magnetite and amphibole, 

side by side in the lower cherty beds 6 

IV. A. The "intermediate" slate of the Biwabik forma- 


B. The spotted fayalite rock in the slaty beds above 

the intermediate slate 8 

V. A. The cherty taconite of the main magnetite beds 

on the east Mesabi 

B. Two close views of the conglomerate 20 

VI. A. Cherty taconite specimen showing the irregular 
bedding and the parallelism of the flat peb- 
bles of the conglomerate 
B. The upper slaty beds, showing some nodules of 
coarser cherty material, grading into lenticu- 
lar beds of similar material 20 

VII. A, B, and C The cherts with structures probably 

due to algae 22 

VIII. A and B. The septaria cracks in coarse cherty nod- 
ules of the upper slaty beds 22 

IX. A, B, C, and D. A series of grains from 0.1 mm. to 

30 mm. showing similar forms, composition, 

and detailed structures 24 

X. A, B, C, D, and E. Drill cores showing the several 

horizons recognized as clearly as the outcrops 26 
XI. A. Drag folds in the upper slaty beds, alternating 
with more competent layers 
B. The closely spaced joints of the conglomerate 

phase of magnetite deposits ; . 34 

XII. Detail map of the Mesaba area 46 

XIII. Detail map of the Spring mine area 48 


XIV. AA'. Cross section of the iron formation east of the 

Spring mine 

BB'. Cross section of the iron formation in sec. 17, 

T. 60 N., R. 12 W 48 

XV. Detail map of the Ridge area 50 

XVI. Detail map of the Iron Lake area 52 

XVII. Detail map of the Sulphur area 52 

XVIII. Detail map of the Birch Lake area 54 


Fig. 1. Sketch map of the west end of Lake Superior showing the 

location of the east Mesabi 1 

2. A section along the strike at the base of the iron forma- 

tion in sec. 17, T. 60 N., R. 12 W., showing the uncon- 
formity and the steep slopes of the granite floor 6 

3. Curve showing the variation in magnetic iron content in 

the several horizons of the iron formation in a well ex- 
plored section on the east Mesabi 16 

4. Correlation chart of the Biwabik iron formation from cen- 

ter to the east of the Mesabi range 24 

5. Sketch map of a quarter section about 20 miles west of 

Mesaba, in which the drilling has developed a body of 
hematite 33 

6. Block diagram of the fold near Spring mine 50 

7. Sketch of a cross section 200 paces south of the north side 

of N.E.J sec. 25, T. 60 N., R. 13 W., showing outcrops 
and (in broken lines) the structures supposed to exist 
below 52 

8. Sketch of a cross section near the S.W. corner sec. 26, T. 

60 N., R. 13 W., showing outcrops and (in broken lines) 

the structure supposed to exist below 53 

9. Cross section of the iron formation south of Birch Lake, 

showing outcrops and dips, which are extended under- 
ground 55 




The Mesabi range is a belt of iron-bearing formation about 100 
miles long, located about 80 miles north of Duluth, which is situated at 
the west end of Lake Superior. The trend of the belt is east-north- 
east. (See Figure 1.) The iron-bearing formation, commonly called 
taconite, is largely drift covered throughout the main range and has 
few of the topographic features of a "range." It is called a range be- 
cause iron-bearing formations in other districts form ranges; and at 
the east end of the Mesabi district there are some rocky hills rising 
200 to 400 feet above the general level. 



This report is a discussion of the eastern end of the range — that 
part which lies between the town of Mesaba and Birch Lake, a distance 
of about 20 miles. This portion of the range is commonly called the 



"east Mesabi." It is distinguished from the rest of the range by sev- 
eral features, besides the topography above mentioned. In this area, 
outcrops are numerous; most of the iron is in magnetic form; recrys- 
tallization has increased the size of grain and reduced the porosity; 
there has been very little leaching or enrichment; and in places the 
beds are more highly tilted. These several peculiarities make the 
east Mesabi a logical unit for separate study. 

The field work was done in the summer of 1917, by Frank F. Grout 
and T. M. Broderick. The Mesabi Syndicate (D. C. Jackling, and 
Hayden-Stone and Company) were at that time actively exploring 
certain parts of the area and Mr. W. G. Swart, in charge at the Duluth 
office, made the work much more effective by rendering many drill 
records accessible to the Survey. A large number of samples have 
been assayed in their laboratories. 


Leith 1 has summarized the literature of the Mesabi range as a whole 
up to 1903. It was the magnetite of the east Mesabi that first 
attracted the attention of explorers. The iron ores are mentioned in 
a report of 1866. 2 An exploration for iron was made in 1875 and cov- 
ered very little except the east Mesabi outcrops. 3 All reports indicated 
that no rich ores were found in large quantities. The later discovery 
of the richer hematite ores in test pits farther west led to the rush of 
explorers in 1891 and 1892. Several explorations are reported, both 
before and after this rush, by N. H. Winchell and H. V. Winchell of 
the Geological and Natural History Survey of Minnesota. 4 Spurr 
also prepared a report on the Mesabi iron-bearing rocks. 5 In 1903 
Leith issued the monographic report of the United States Geological 
Survey on the district, 6 and a later monograph on the Lake Superior 
region includes additional discussions of the Mesabi district. 7 Since 
the discovery of hematites the east Mesabi has been relatively 
neglected. Two small hematite mines east of Mesaba have been 
exploited and for the most part worked out, and no others have been 

1 Leith, C. K., The Mesabi iron-bearing district of Minnesota: U. S. Geol. Survey Mon. 43, 1903. 

2 Eames, Henry H., The metalliferous region bordering on Lake Superior: First Report of the 
State Geologist of Minnesota, St. Paul, 1866. 

1 Chester, A. H., The iron region of northern Minnesota: Geol. and Nat. Hist, of Minn. Eleventh 
Ann. Rept., pp. 154-67, 1884. 

4 See for example, Iron ores of Minnesota: Geol. and Nat. Hist. Survey of Minn. Bull. No. 6 
by N. H. Winchell and H. V. Winchell, 1891. 

1 Spurr, J. E., The iron-bearing rocks of the Mesabi range in Minnesota: Geol. and Nat. Hist. 
Survey of Minn. Bull. 10, 1894. 

• Op. cit. 

7 Van Hise, C. R., and Leith, C. K., Geology of the Lake Superior region: U. S. Geol. Survey 
Mon. 52. 




In the logical continuation of the work of Leith on the Mesabi range 
published in Monograph 43 and part of Monograph 52 of the United 
States Geological Survey, it is to be expected that the detailed petrog- 
raphy of the several parts of the iron formation will be the basis of a 
more detailed stratigraphy. The beginnings of this subdivision of the 
iron-bearing formation have been made by Leith himself in notes on 
certain persistent horizons and in maps showing a persistent slate bed. 
More in detail are the sections worked out by the engineers of the 
Oliver Iron Mining Company — Kingston, Wolff, Cronk, and others — 
working under J. U. Sebenius. The sections published by Wolff 8 show 
numerous persistent beds with recognizable peculiarities. 

The east Mesabi is particularly suitable for a study of detailed 
stratigraphy, because outcrops are vastly more numerous than in other 
parts of the range, and there are also many drill cores available. The 
season of 1917 was especially favorable for mapping on account of the 
recent forest fires and lack of vegetation. It is now thought that the 
less developed eastern end of the range reveals several recognizable 
horizons which will aid one to determine the structure and some of the 
ore reserves in other parts of the range. 

The immediate incentive to the study of the east Mesabi was the 
certainty of magnetite bodies of a large size, and the prospect that they 
were of a quality and so located as to warrant magnetic concentration. 
This is a commercial affair and the continued success of such a venture 
is a matter of close calculation. The costs of several processes involved 
have been much reduced, and the prices of good low-phosphorus ore are 
such as to encourage an attempt to mine these ores. In any event, 
with the reduction of the hematite reserves, interest in the leaner con- 
centrating ore will increase. The details of the occurrence of the mag- 
netite will be of interest in the future, even if present exploration is 
not continued. 


In a sedimentary rock, such as the taconite, magnetite may accu- 
mulate in several ways: 

1. By mechanical concentration of magnetite in sand. 

2. By deposition as ferric oxide and subsequent alteration to mag- 

3. By deposition as ferruginous chert in a I eaner condition and 
enrichment to ferric oxide, as ores are now being enriched; and later 
by alteration to magnetite. 

8 Wolff, J. F., Recent geologic developments on the Mesabi iron range: Amer. Inst. Min. Eng. 
Trans., vol. 56, p. 148, 1917. 



4. By deposition as magnetite from solutions emanating from 
gabbro or granite near the contact. 

The determination of the actual origin, then, involves a determi- 
nation of the original mineral and textural character; the time the 
iron oxide accumulated; whether it was enriched, and if so, when; 
what processes altered the ferric oxides to magnetite; and what igneous 
rocks contributed to it or altered it. The first drilling that was done 
ignored the question of origin and only attempted to determine that 
the known outcrops of ore were connected by similar material under 
cover, and the depth to which the ore extended. When the matter 
of origin was considered, a small amount of drilling soon confirmed 
the observations of outcrops, indicating that (1) the concentration 
was primary, (2) the original oxides, probably mostly limonite, were 
altered by general metamorphism to magnetite, (3) none of the igneous 
rocks have essentially modified the iron content of the taconite, 
(4) most of the beds showing a concentration of iron are characterized 
by a conglomerate texture. 




Cray slate with no secondary cleavage. 





Thin bedded (aconite with thin seams 
of magnetite becoming abundant 


. - 



Presence of septaria. drag folds, brec- 

magnetite contents. Garnet developed 
commonly toward bottom. 

18 i 







32 i 



with conglomerate texture alternate 
with beds of magnetite. The matrix 
of the conglomerate pebbles is gran- 
ular quart?, magnetite and amphibol;; 



to slaty texture. 


Au b. 



Bbck state, easily broken into thin 




-Oj« M jagc «r«c IU r« «««blw 


Magnetite rich lone with some crap., 
ulcs and conglomerate, but in many 
places a later breceia.ion. In places 



wackc and slaty phases. Shows cross 

torn with boulders and pebbles, many 
of which arc from the underlying 



Pink, porphyritic, moderately quart- 





All the formations of the east Mesabi, as shown in the following 
table, are in the Algonkian system except the thin Pleistocene deposits. 
For detailed descriptions of the several formations, reference should 
be made to the monographs of the United States Geological Survey 
concerning the region as a whole. 1 Brief notes are given below, but 
some details are added in the case of the iron formation. The map 
(in pocket) shows their areal distribution. 

Quaternary system 

Pleistocene series Deposits of late Wisconsin age 

Algonkian system 

Keweenawan series Duluth gabbro and diabase dikes and sills 

Unconformity (only intrusive contacts in this area) 
Huronian series 

[Virginia slate 

Upper Huronian < Biwabik formation (iron-bearing) 

[ Pokegama quartzite 


f Giants Range granite 
Lower-Middle Huronian s 

[Slate and schist formation 


The slate and schist formation. — These slates and schists are light 
green rocks of variable grain. Most of them show a nearly vertical 
cleavage and bedding which strike a little north of east. They include 
several phases, not very extensive, which apparently represent old 
intrusions and beds of different character. They extend from the rail- 
road at Mesaba to a point east of the Spring mine in R. 14 W. 

The Giants Range granite. — The granite is in general pink, porphy- 
ritic, moderately quartzose, hornblende granite. It was intruded 
north of the present outcrops of green schist and forms the main ridge 
behind them. The contacts show the intrusive nature of the granite. 
In much of the east Mesabi the schists were eroded before Upper 
Huronian time and the iron formation rests on the granite. These 
contacts do not show any intrusions of granite into the sediments. 
The granite had evidently been weathered before the sediments were 

1 Van Hise, C. R., and Leith, C. K., The geology of the Lake Superior region: TJ. S. Geol. Survey 
Mon. 52, 1911. 

Leith, C. K., The Mesabi iron-bearing district of Minnesota: U. S. Geol. Survey Mon. 43, 1903. 









deposited, for it now shows an upper zone (and a zone around many 
pebbles in the conglomerate) altered to a quartz-chlorite-garnet rock. 

Unconformity. — The erosion preceding the deposition of the Upper 
Huronian seems to have cut deeply into the slates and granite, and 
although the general trend of the contact with the over-lying beds is 
regular, there are minor irregularities indicating a relief of 50 to 100 
feet in many valleys in the pre-Animikie surface. Since the sediments 
are more easily eroded than the basement complex, the present valleys 
tend to follow, in many cases, the older valleys. (See Figure 2.) 


The Pokegama quartzite. — The Upper Huronian (Animikie) forma- 
tions lie on the older beds with. a conspicuous unconformity in most 
places. The boulders of slate or granite in many places show the der- 
ivation from the slate or granite below. Above the coarse conglom- 
erate, and in some cases directly upon the older rock, are finer sedi- 
ments of several kinds. Where the sediment is quartzite it is given 
the name of Pokegama as a formation. Where it is cherty and ferrugi- 
nous it is the Biwabik formation. In many places there is a mixture 
of clastic grains and a good deal of magnetite, leaving some doubt as 
to its classification. 

The Pokegama quartzite is a hard, pink quartzite in lenses with a 
maximum thickness of 30 feet in the eastern district. Where it is prom- 
inently developed several exposures show a conglomeratic phase at the 
top. The conglomerate pebbles in most cases are not fragments of 
the Pokegama, such as would indicate its erosion; but they are chert 
fragments, probably indicating a breaking up of newly formed sedi- 
ments above the sand. 

The Biwabik iron-bearing formation. — The conglomerate, which is 
above the granite and schists in most places along the east Mesabi, 
grades into a series of ferruginous cherty rocks of unusual and varia- 
ble character. These make a formation 350 to 500 feet thick. Its 
mineralogy, petrography, and structure are described in detail in later 

The Virginia slate. — The Virginia slate lies with apparent con- 
formity upon the Biwabik iron formation, but the change in material 
is somewhat abrupt. The cherty ferruginous iron formation may be 
thin bedded (slaty in texture) in certain parts, but the Virginia slate 
is slaty in both texture and composition. The contact of the forma- 
tions is marked by a layer of limey carbonate, rather than a gradation 
from chert to slate. 

The total thickness of the slate can not be measured in the eastern 
district, for its upper part is everywhere intruded by the Duluth 



gabbro. The gabbro lies above several hundred feet of slate near 
Mesaba, but cuts down gradually across the slate into the iron forma- 
tion near Dunka River, in R. 12 W. 

Tlie sills and dikes (of uncertain age) . — Intruded into the iron forma- 
tion and the lower part of the Virginia slate are several sills 10 to 20 
feet thick. They are much harder than the slate and form conspicuous 
ridges. Drilling has revealed also several dikes of similar material. 
The diabase of the sills is characterized by an abundance of coarse 
white feldspar phenocrysts in most exposures. Each good outcrop shows 
that the phenocrysts have concentrated in the upper half of the sill. 
While the sills are mostly fine grained, at many places the phenocrysts 
are as much as 2 inches in diameter. 

The diabase of the sills and dikes would naturally be correlated with 
the gabbro which is of similar chemical composition, especially since 
most of the exposures occur at the east end, very close to the great 
gabbro area. It would be logical to consider them apophyses of the 
gabbro, except that detailed study shows that they are metamorphosed 
by contact action of the gabbro. Hence, although they are later than 
the slate and earlier than the gabbro, it is uncertain whether they were 
intruded in Upper Huronian or early Keweenawan time. Igneous 
action was more characteristic of Keweenawan time and in the table 
of formations given above the intrusions are placed in this division. 


The Duluth gabbro. — Along the eastern Mesabi practically all of 
the gabbro is a coarse olivine gabbro with a somewhat variable compo- 
sition, in alternating bands. The bands dip eastward at a steep, but 
variable angle. The texture is coarse up to the contact, though in 
some places the contact is determined only by close study. 


The intrusions of diabase and gabbro have produced alterations in 
the earlier rocks. The effects of the diabase are not conspicuous except 
where it is in contact with slates, which are crumpled and recrystallized 
in a thin zone. The rock is greatly hardened, and spots of cordierite 
and amphibole appear. A chilled border phase of the diabase is also 
spotted, containing spherulites, and may be confused with the cor- 
dierite rock in the field. 

The granite of this area has been described by Leith 2 as intrusive, 
into the upper Huronian, but it has produced no visible contact effect, 
and is now believed to be older than the sediments, as it was mapped 
by Winchell. 3 

1 Leith, C. K., The Mesabi iron-bearing district of Minnesota: U. S. Geol. Survey Mon. 43, 1903. 
1 Winchell. N. H., Geol. and Nat. Hist. Survey of Minnesota Final Report, vol. 4, plate 67, 1899. 





I The gabbro is the largest mass of igneous rock in the region and 
one of the largest intrusions in the world. The contact effects extend 
in some places several hundred feet. Probably the most important 
effect is a very complete recrystallization of practically all the adja- 
cent formations, into a sugary textured hornfels. The weathered sur- 
face of the contact rock is at places much like brown sugar in appear- 
ance. For this reason, it was referred to as "muscovadite" by Win- 
chell. 4 The uniformity in appearance of the hornfels, whether from 
slate, diabase, or even from some iron-bearing formation, leads natur- 
ally to the conclusion that contributions from the gabbro have con- 
siderably changed the composition as well as the texture of the con- 
tact rocks. This, however, has not been proved, and the beds are so 
recrystallized that it is difficult to correlate altered and unaltered beds, 
if it was desired to make a comparison. 

The iron-bearing formation contains bands of quartz and magnet- 
ite, which have resisted the apparently general tendency to become 
hornfels. Recrystallization develops coarse quartz grains up to half an 
inch thick from chert, amphiboles several inches long from the cherty 
carbonates, and magnetite in large crystals from pebbles and dust of 
iron-bearing minerals. It might be expected that iron oxide in the 
layers would react, during metamorphism, with quartz layers, pro- 
ducing silicates, but there is little evidence of such action. On the 
other hand, white quartz bands are much more conspicuous than the 
original gray chert bands alternating with iron oxides, and it is pos- 
sible that during recrystallization some disseminated magnetite in 
chert has been more or less segregated into bands. One thin section 
made from pyroxenite, formed by contact action of the gabbro on the 
taconite, shows clear evidence of replacement by iron oxide along the 
borders and cracks. The temperature during recrystallization may 
have reached 575° C. or more, 5 near the gabbro. 

In places, where the gabbro is close to the several horizons of iron 
formation, pegmatitic stringers are visible in great numbers. The 
feldspars are coarse and there is a great deal of graphic intergrowth 
with quartz. Farther away similar patches appear but are less defi- 
nitely pegmatitic — feldspar is not so prominent, but quartz stringers 
with vugs are found in similar positions. These stringers are not 
connected by visible dikes or cracks, but suggest a pervasive emanation 
from the gabbro. No sign of so much feldspathic material is visible in 
the less altered iron formation. To check the matter, however, samples 
were taken from the same horizon near the gabbro and farther away. 

* Winchell, Alexander, Geol. and Nat. Hist. Survey of Minn. Fifteenth Ann. Rept., p. 183, 1886. 
IMuscovado" was the early settlers' name for brown sugar. 

* Van Hise, C. R., and Leith, C. K., op. cit., p. 549. 




The alkalies were .78 per cent near the gabbro and .38 per cent at a 
distance. The difference, equivalent to about 3 per cent feldspar, may 
possibly give a fair idea of the additions of alkalies that took place. 

The long time and many events recorded in the geologic history of 
these formations make it difficult to determine how much of the rock 
alteration is due to contact action and how much to regional meta- 
morphism. The progressive increase in coarseness of grain toward 
the gabbro indicates that this coarseness is largely a contact effect. 
However, the burial of the iron formation under a great thickness of 
slate favored some general metamorphism. This is shown by the indu- 
ration and partial recrystallization of the slate itself, and by the drag 
folds in the less competent beds of the iron formation. 

Some of the mineralogical changes in the iron formation seem to 
vary with the amount of recrystallization and the proximity of the 
gabbro; others are apparently independent of position and may be 
attributed to regional metamorphism. For example, it seems unlikely 
that the original deposit contained iron in a magnetic form, but at 
present the central beds of the iron formation, where unaffected by 
weathering, contain about 30 per cent of magnetically separable iron 
oxides, all the way from the gabbro at the east end of the range, to 
Coleraine, 60 miles from any gabbro exposures. This may result from 
a reduction of primary ferric oxides, either by simple heat, by ferrpus 
compounds, or by organic material which is now represented by 
graphitic beds. There are, in many specimens, intimate mixtures of 
oxides, which give a red streak but are attracted by an ordinary magnet. 


The general structure is simple, the beds dipping south and south- 
east at angles of about 5 degrees, on the side of the great Lake Superior 
basin. Locally, disturbances such as the gabbro intrusion produced 
steeper dips. The drag folds, a few feet in extent (Plate XI A) in the 
thin bedded formation are so placed that they may be related to this 
larger structure. However, there are folds of intermediate size, one 
hundred to several hundred feet across. Most of the irregularities 
shown on the map in the boundaries of the several formations, are due 
to the valleys crossing gently dipping beds. However, the folds like 
that at the Spring mine, or south of Iron Lake, are more likely related 
to some set of forces other than those which produced the major struc- 
ture, possibly to igneous intrusions or some local modification of the 
larger stresses by the bed rocks below. The intrusion of the gabbro is 
very likely responsible for the sharp turn northeast of Dunka River. 
A few small domes in the lower cherty beds are probably due to the 
irregularity of the floor on which the beds were deposited. 



No faults were noted in the iron formation. If present they are 
probably of very local significance. 


Throughout a large portion of northeastern Minnesota, glacial ice 
has scoured off the products of weathering and left relatively thin 
deposits of gravel. This seems to be true of the east Mesabi range, but 
on the main part of the range soft hematites have not been entirely 
removed, and a deeper deposit of drift covers them. The east end of 
the range stands at a level perhaps 200 feet above the west end, and 
it is a question whether the absence of soft ores at the east end is due 
to the fact that they never were formed, or to their formation at a 
higher level more exposed to glacial scouring, or to a difference in their 
relation to the lobe of ice that moved across them. Van Hise suggests 
that some of the range occupies an interlobate position. 6 Many of the 
rock surfaces are polished and scratched as if considerably glaciated; 
but in the rock of these high lands there are some gorges cut about 
100 feet deep, which have not been obliterated by the glaciation, and 
it does not seem likely they were formed since glaciation. Possibly 
the soft ores are absent for the same reason that the rocks stand higher 
than those farther west; not because of glaciation, but because the 
rocks were recrystallized, hard, and resistant to weathering. 


The effects of weathering on the iron formation are very slight in 
the east part of the Mesabi range. The cherts stand out as ridges on 
the slope below the main granite ridge. The Virginia slate forms a 
lowland between the gabbro on the south and the granite and iron 
formation on the north. The slaty layer below the middle of the iron 
formation is also marked in many places by a valley. These topo- 
graphic effects are probably due to the more rapid weathering of the slate. 

In exposed places the cherty iron ore has stood under the weather 
probably for centuries without apparent attack. On the other hand, 
ground waters of different kinds produce some effects. The leaching 
of silica at the hematite mines has probably resulted from alkaline 
solutions which might originate near the granite. Contrasted with 
this, the swamp waters are acid and any taconite in the water loses 
its iron minerals; magnetite and fayalite dissolve rapidly, amphiboles 
next, and quartz remains. Most of the springs issuing from the iron 
formation and the waters reached in drilling, are almost free from 
iron. In swampy areas there are some limonite deposits. A spring 
carrying considerable iron issues in sec. 27, T. 60 N., R. 13 W. 

•Van Hise, C. R„ Iron-ore deposits of the Lake Superior region: U. S. Geol. Survey Twenty- 
first Ann. Rept., pt. 3, pp. 411-12. 



The following minerals have been noted in the iron formation of the 
east Mesabi. 

Quartz is the most abundant mineral in the recrystallized taconite. 
It occurs also in fragmental grains and a few small veins and septaria 
cracks. Chert and jasper make up some large primary beds, two of 
which show structures indicating organic (probably algal) origin. Some 
of the cherts and jasper layers were broken into fragments and pebbles. 

Amphibole is the second mineral in abundance in the main taconite 
beds. It was probably formed by metamorphism of the greenalite and 
other silicates, that may have been present in the primary deposit. 
It is now intergrown with the quartz and magnetite. Rounded 
markings, consisting of amphibole in quartz, are seen in some thin sec- 
tions — apparently the only remaining sign of an original granule tex- 
ture. Some conspicuous grains, poikilitically enclosing quartz, etc., 
have grown to be 6 inches or more in length. Dark amphiboles are 
conspicuous also in pegmatitic stringers near Dunka River. Most of 
the amphibole is light green, and may be classed as actinolite, but 
other specimens vary from yellow to black. Grunerite and cumming- 
tonite have been reported. The darker silicate in many specimens is 
probably grunerite. 

Magnetite forms variable layers almost pure, up to 6 inches or more 
thick, alternating with the more common intergrowth of quartz, amphi- 
bole, and magnetite. Magnetite occurs also as pebbles, small rounded 
granules, crystals, and most commonly intergrown with quartz and 
amphibole, making up small rounded granules, and the matrix of the 
granules. There is a slight concentration of magnetite in a zone at the 
border of many granules. Recrystallization is so complete that little 
trace of granules is visible except these rings of fine grained magnetite. 
No attempt has been made to distinguish magnetite from other mag- 
netic iron oxides. Ordinarily it is probably correct to call a magnetic 
oxide with a black streak, magnetite. In some parts of the Mesabi 
range there may be other magnetic oxides. 

Hematite and limonite were rarely seen in the east half of the area 
except as a slight film of weathered mineral. At several places in the 
western part the leaching and oxidation have produced considerable 
bodies of these ores. In much of the western belt, the fresh black mag- 
netic ores also show in places a red streak. The hematite and mag- 
netite are intimately mixed, and magnetic concentration separates 
most of the iron oxides from the silicates. 



Garnet is developed at several horizons in the iron formation. In 
the chloritic alteration product of the weathered granite and con- 
glomerate at the base of the formation are many small, well formed 
red crystals. A more massive brown garnet forms layers and spots 
with the amphiboles, etc., near the slaty beds of the formation. While 
garnets may be found at a number of horizons, the only other promi- 
nent occurrence is above the main conglomerate beds; here also it is 
brown and in irregular spots. 

Fayalite is conspicuous in crystals up to an inch in diameter. Most 
of them are about the size of peas. The fayalite is yellow when fresh, 
and brown when weathered. Thin sections show that most grains have 
thousands of microscopic inclusions of quartz and amphibole. Some 
have also magnetite. Fayalite was seen in greatest abundance in the 
slaty taconite above the main slate of the iron formation, and was 
best developed in R. 12 W. 

Pyroxenes include augite, diopside, and hypersthene. The augite 
forms metacrysts resembling fayalite, in the recrystallized cherty taco- 
nite. Hypersthene was seen only in pegmatitic stringers near Dunka 
River. Babingtonite has been reported. 

Mica occurs as microscopic grains in the altered slaty horizons. A 
larger specimen was seen in a drill core from the taconite near Sul- 

Feldspars are not seen in the main outcrops, but both orthoclase 
and plagioclase occur in the pebbles of granite in the basal conglom- 
erate; also in the pegmatitic stringers near the gabbro. 

Carbonates are to be found in small amounts in the ore horizons 
intergrown with quartz, amphibole, and magnetite. A layer rich in 
calcium and possibly other carbonates is more or less continuous at 
the upper contact of the iron formation, and some carbonate appears in 
the lean taconite, just below this layer, — Aub 6 . The magnetite-rich 
bed below the intermediate slate also contains carbonate. Secondary 
carbonates have been deposited in openings in the leached zones, where 
there has been enrichment, and in pseudomorphs after some of the 
granular greenalite. 

Greenalite has been found mostly in the more western parts of the 
area. Nearer the gabbro, recrystallization has altered it to several 
other minerals, and more or less modified the texture of the rock, so 
that little greenalite is left. We have checked the low alkali content 
of greenalite rock by several analyses. The granules of amphibole may 
form from greenalite with little change in material. 

Graphite. — Many black layers and flint "concretions and pebbles in 
the midst of magnetite-quartz masses prove to contain little or no 



iron. The solution of the silica and iron by successive treatments with 
hydrofluoric and hydrochloric acids leaves a soft black residue, appar- 
ently graphite. 

Apatite is not visible in most thin sections, but has been detected in 
the coarser recrystallized ore at the east end of the Mesabi range. 

Chlorite is characteristic of the basal conglomerate where the feld- 
spars were weathered before metamorphism. Some is believed to 
occur also in the aluminous beds higher in the formation. 

Pyrite occurs as numerous scattered small bunches below the slate 
member of the iron formation, and is scattered in small amounts at 
other horizons. 

Pyrrhotite, chalcopyrite, and arsenopyrite have been noted in drill cores. 
Epidote and tourmaline are visible in the contact zone near the 

Kaolin may be present in some of the slates. 
Malachite is derived from chalcopyrite near the surface. 


It has become the custom on the Mesabi range to refer to practi- 
cally any phase of iron formation (except hematite ore and paint rock) 
as taconite. Most of the rocks are derived from an original sediment 
containing granules that may have been greenalite, but the greenalite 
is so altered in most specimens that the name greenalite-rock is applic- 
able to very few. There are, of course, small amounts of some ordi- 
nary sediments, conglomerates, quartzite, and slate, with the usual 
range of impurities, but these make up. a very small proportion of the 
whole. From the prominence of fine quartz, the rocks may properly 
be classed as cherts, with various qualifications to distinguish the min- 
erals intergrown with the quartz; thus there are ferruginous cherts, 
amphibolitic cherts, calcareous cherts, and sideritic cherts. These 
so-called cherts, however, differ considerably in origin from the con- 
cretionary and precipitated cherts of limestones and common sedi- 
ments. They are derived by alteration from the rocks with granules 
like those of greenalite. Furthermore, in certain places, there are, in 
the iron formation, chert concretions of a more normal sort, and it is 
somewhat confusing to make no distinction. For the rocks of the east 
Mesabi there are several reasons why the name chert is not very satis- 
factory. The granules, analogous to the greenalite structures farther 
west, vary in size up to pebbles 6 inches in diameter. Recrystalliza- 
tion has enlarged the grains so that many specimens no longer have 
any resemblance to common chert. For these reasons we use the local 
name, taconite, in its broader sense for any part of the iron forma- 
tion. Certain phases may be easily classified as chert, amphibolite, 



pyroxenite, quartzite, amphibole-magnetite rock, or magnetite; others 
have so many rare minerals or such unusual proportions of minerals 
that no name has been given and a descriptive name becomes very 
cumbersome. Added to this, it may be necessary to use a textural 
qualification. There are, for example, some cherty amphibolites with 
fayalite metacrysts; and some conglomeratic augite-amphibole-quartz- 
magnetite rocks. Probably no elaborate additions to the nomenclature 
are desirable, and the general term taconite may be understood to cover 
these recrystallized rocks. 


The several divisions of the iron-bearing formation here tabulated 
(Plate III) are based to a large extent on the appearance of outcrops. 
Wolff 1 has emphasized the fact that the several grades of Mesabi ore 
are more or less related to the beds from which they were formed; and 
this relation is even more pronounced in the magnetite than in the 
hematite ore. The divisions previously made by Wolff are indicated 


Virginia slate 

| V Per cent Iron Hi magnetic form 

Carbonate zone 

Lean laconlte 

10 20 30 

Thin bedded uconite 

Septaria, thin beds, 
co'icrenons. etc 


1 ' 


Riclvconglomeratc beds 


Medium conglomerate 


Chenv taconitt 
Little conglomerate 
Slaty beds 


Intermediate slate 

Alga] chert 

Magnetite uconite 


Pokegama quaruite 



» Wolff, J. F., op. cit. 



in the diagrammatic sections (Figure 4), and were useful in the struc- 
tural studies on the east Mesabi; but the thickness of the main divi- 
sions is very different in this part of the range from those plotted by 
Wolff. Some of the divisions are given detailed descriptions on pages 
19 to 26. 

Tabular Section of the Biwabik Formation Where Prospected for Iron on the Eastern 


Top Virginia Slate 

Upper Slaty Beds 

Aub6 Limy carbonate layers. Less than 5 per cent magnetite 5 to 10 feet 

Banded amphibole and white quartz in thin layers up to 6 

inches thick. Less than 5 per cent magnetite 40 to 50 

Aubg Taconite in thin beds mostly less than }/% inch thick. About 

20 per cent iron in magnetite, decreasing at the top. ... 25 to 35 
Taconite in thin beds, like above, but alternating with 
thicker gray beds and concretions with a granule texture, 
and white quartz septaria. The thin beds are drag fold- 
ed and even brecciated. A zone of garnets occurs near 
the bottom. About 20 per cent iron in magnetite 40 to 45 

Upper Cherty Beds 
Aub4 Dark heavy taconite, with some conglomerate texture, and 
granules, alternating in thick beds with thick magnetite 
layers. About 30 per cent iron in magnetite 10 

Jasper and chert with algal structure and conglomerate. 

About 15 per cent iron in magnetite 1 to 10 

Dark heavy taconite like the bed above the jasper. About 

30 per cent iron in magnetite 20 to 30 

Gray taconite with conglomerate and granule texture and 
many thinner lenticular beds of magnetite. About 20 
per cent iron in magnetite 80 to 110 

Lower Slaty Beds 
Aub3 Fine massive to slaty quartz amphibolite with only obscure 
granule structure, and few magnetite beds. Fayalite 
crystals in places. About 10 per cent iron in magnetite . 65 
Aub2 • Black thin bedded slate, more or less recrystallized. About 

5 per cent iron in magnetite 25 

Lower Cherty Beds 
Audi White to gray chert with coarse algal structures. Less 

than 5 per cent iron in magnetite 10 to 15 

Variable taconite with some cherts, breccias, fragmental 
sands, garnets, etc. From 25 to 35 per cent iron in mag- 
netite 2 to 52 

Basal beds; conglomerate in many places, and (in absence 
of Pokegama quartzite) some green shales. Less than 
5 per cent iron in magnetite to 15 

350 to 470 

Base Giants Range Granite; or Lower-Middle Huronian 

Slate; or Pokegama Quartzite. 

1 8 



The bottom of the lower cherty beds, Aubi, is for the most part a 
conglomerate of irregular thickness, which fills in some of the erosion 
valleys in the older surface. It is not separately shown on the map. 
Where it rests on granite, as it does in most of the east Mesabi, some 
of the boulders are clearly of the same granite. Evidently the weather- 
ing of the boulders and the granite below left an aggregate which under 
later metamorphism developed chlorite and garnet, between the orig- 
inal quartz grains that seem to remain unaffected (Plate III A). 
Between the pebbles and in thin beds overlying them are greenish 
shaly beds, probably derived from the same weathered granite. Where 
the iron fo mation rests upon the older slates or the Pokegama quart- 
zite, this basal bed consists mainly of quartz and chert pebbles. 

Above these basal beds is the lowest of the magnetite beds. It is 
not recognized at the east end of the area, but is found a few feet thick 
in R. 13 W., and over 50 feet thick near Mesaba. It contains magnet- 
ite layers more or less mixed with sand and cherts. (See Plate III C.) 
There are phases that resemble the conglomerate of the upper cherty 
beds, but more that seem to be brecciated after deposition. 

The increase in thickness toward Mesaba and the high per cent 
of iron make this bed of interest as a prospective ore horizon farther 
west. It is no doubt the bed which Wolff describes as producing "hard 
blue ore" in the "brown cherty taconite. ,,2 Toward the east where it 
is thin it was missed in so many places that it has not been mapped as 
a continuous belt. 

At the top of the lower cherty beds, a hard chert crops out in many 
places, indicating a continuous formation over the length of the range. 
Most of the rock is of fine grain and has a peculiar structure, empha- 
sized in gray and white or jaspery bands. (See Plate III B.) The bed 
is not so lenticular and irregular as the concretionary beds, such as 
occur above the slate, but is characterized by a more regular waviness, 
probably due to organic action. It closely resembles several forms 
from the pre-Cambrian ascribed to algal growth. Similar forms were 
recently described and illustrated by Moore from the iron formations 
near Hudson's Bay. 3 These forms on the Mesabi are almost wholly 
chert, while the others are characteristically calcareous. Nevertheless 
they are considered of organic origin, for it is likely that many early 
forms of chert are analogous to later calcareous forms. 4 The concentric 
masses are 6 to 18 inches across. (See also the description of smaller 

» Op. ext., p. 154. 

1 Moore. E. S., The iron formation on Belcher Islands: Jour, of Geology, vol. 26, pp. 425-26, 1918. 
« See Clarke, F. W., Geoehemical evidence as to early forms of life: Jour. Wash. Acad. Set., vol. 
6. p. 603. 1916. 



algal forms in the upper cherty beds, page 22.) The gray color of 
some bands may be due to their magnetite content, but only a few 
thin layers in v the cherts have a high per cent of iron. In some out- 
crops the banded cherts include fragments of iron oxide, so that the 
rock greatly resembles the conglomerate beds of the higher divisions 
of the formation. The thickness of these lean beds below the slate is 
about 15 feet and is fairly constant in the eastern area. 


The lower part of the lower slaty beds, i\.ub 2 , is the most notably 
slaty in structure of any part of the iron formation. It is known 
through the main range as the "intermediate slate." (See Plate IV A.) 
In contrast with the lower beds it is very constant in thickness and 
bedding through the whole range. It is probable that the lower beds 
had filled up the irregularities in the bed rock surface before the time 
of deposition of the slate. It is so much more like an ordinary shale 
that it is natural to attribute the material to normal mechanical depo- 
sition and, in contrast, to attribute the ferruginous cherts to precipi- 
tation. Well-developed ripple marks are seen in places. None of 
these slaty beds contains much magnetite. Extensive recrystalliza- 
tion at the east end does not prevent the weathering to thin slabs along 
the bedding, at most outcrops. An analysis of the slate by George 
Steiger is reported by Van Hise and Leith. 5 

This bed grades into overlying rock of a very different general char- 
acter, with no sharp break. The slates, about 25 feet above the bottom 
of the bed, become more granular in texture and show less regular 
bedding. Higher up the rocks are hard and fine grained, with con- 
choidal fracture. They are more massive in structure as compared 
with the thin beds below. Thin sections show a mat of very fine 
amphibole needles in chert, with a very obscure granule structure. 
In a few places small areas of conglomeratic texture appear, with small 
magnetite pebbles, but abundant magnetite is found in few beds, and 
as a whole, in very small amounts. In the area near Sulphur the rocks 
of this division show fayalite in many outcrops. Plate IV B shows the 
spots that develop as the. fayalite weathers brown. As a division of 
the formation it will be called the siliceous bed above the intermediate 


The lower part of the upper cherty beds is the lowest bed with a 
thickness as much as 50 feet that contains a large per cent of magnet- 
ite. The line between this bed and the more siliceous beds below is 
not sharp, but the formation in a few feet becomes granular, magnetite 

« Op. «/., p. 191. 



pebbles grow more numerous, and layers of pure magnetite become 
thicker and more abundant. The whole mass is characterized by the 
occurrence of the pebbles. No exposure 5 feet thick lacks pebbles. 
Any drill core from these beds that is 5 feet long will show cross sec- 
tions of pebbles. 

A vertical section of the bedding shows it to be very irregular. 
Most of the beds, whether of magnetite, conglomerate, or amphibole, 
pinch out in irregular lenses. Such a bedding is illustrated in Plate V A. 
It is not regular enough for ripple marks; neither is it evidently 
related to folding; both ripples and folding are observed in the beds 
independently of this structure. There are no signs of cross bedding. 
Three processes are suggested below (in discussing the origin of the 
magnetite deposits), which may have made the beds irregular: (1) the 
shrinkage of a colloidal precipitate; (2) the solution of deposited layers; 
and (3) concretionary rearrangement. 

Many of the pebbles of the conglomerate are of magnetite, many 
are of material resembling the beds just below, and many show a core 
of siliceous granular rock with a zone of magnetite around the border. 
Plate V B indicates the variety of pebbles in a fair sample. The shapes 
are mostly somewhat flattened and fairly well rounded, but a con- 
siderable number are angular, and plate-like. Some vertical sections 
of the beds indicate that magnetite layers, which can be traced some 
distance, may be broken up into flat, vertical-sided blocks at one end. 
(See Plate V A and Plate VI A.) 

The matrix between the pebbles in the lower part of this bed is 
much the same as the main portion of the bed under it ; but in the mid- 
dle and upper parts, it is of sandy or granule texture, containing much 
more magnetite, and showing under the microscope structures that 
have been attributed to greenalite. In the altered grains of these 
rocks, there seems to be no difference in the nature of the material in 
the granules and in the coarser pebbles. Grains of the same general 
character range in size from Tt inch up to 6 inches. There is no 
break in the series. (See Plate IX.) 

The magnetite content of these beds is fairly constant if taken in 
large masses. However, it is possible to select five-foot drill cores, or 
five-foot exposures, a few feet apart along the beds, which will vary 
as much as 50 per cent of the total. This is exceptional and the aver- 
age magnetic iron 6 content is probably well above 20 per cent. Total 
iron is about 37.5 per cent. 

The next division of the upper cherty beds is not essentially dif- 
ferent from the main conglomerate, except that the magnetite layers 

* The term "magnetic iron" is used in this report to designate the iron contained in minerals 
which are attracted by an ordinary hand magnet. 







are thicker and more numerous. This seems to be the best magnetite 
deposit as much as 20 feet in thickness. To the west of this area the 
richer beds show a red streak indicating considerable hematite, though 
magnetic concentration separates most of the iron. 

Above the best magnetite deposit is a thin bed of cherty or jasper y 
material with the characteristic algal structure mentioned in the de- 
scription of the upper part of the lower cherty beds. This bed is very 
hard, and so placed in the midst of hard beds that it outcrops in many 
places. However, it is too thin to show on the map. The color varies 
from gray to red, with variation in composition, but the cause of the 
variation is undetermined. The magnetite content of the algal bodies 
is slight (see Plate VII), but between the finger-like masses are some 
pebbles of magnetite and granular material much like the conglom- 
erates in adjacent beds. 

The bodies whose origin is here attributed to algal growth resem- 
ble little piles of thimbles or inverted bowls. They are half an inch to 
three fourths of an inch in diameter, piled in irregular columns about 
six to twelve inches high. The lower parts, like the sides and upper 
parts, seem to merge into the granular fragment al material of the con- 
glomerate. Along the east Mesabi the curving lines of the structure 
do not seem to include any fragmental grains or granule markings 
except along the margins. However, some drill cores from taconite 
near Eveleth show a more confused mixture of banded jasper and peb- 
bles, indicating that the growth could occur simultaneously with frag- 
mental deposition and even concentrically around the grains. Hori- 
zontal sections of these masses show the concentric rings of a concre- 
tion, but the vertical sections are not easily explained as concretionary, 
or even as results of diffusion. It seems more likely that the finger- 
like masses grew, convex surfaces upward, all over the bottom of the 
water. Plate VII shows the characteristic forms. It is noteworthy 
that these small growths, forming a layer a few inches thick, are so 
widespread that, the structure serves as a horizon marker in the drill 
cores examined from places all along the range. The outcrops indi- 
cate a continuous bed 15 miles long and the curving bands are so small 
that they are easily recognized in drill cores far to the west of the outcrops. 

Specimens of the algal material were examined by Dr. Charles D. 
Walcott, who writes; "I am still at a loss to explain the origin of such 
structure without the influence of some organic agency acting in con- 
junction with diffusion and concretionary phenomena. The structure 
is very much like that which occurs in the siliceous limestones of the 
pre-Cambrian in the Grand Canyon, Arizona, and somewhat like those 
found in the pre-Cambrian Belt series of Montana. The Grand 



Canyon forms are referred to Collenia." 7 No cell structures have been 
detected in the thin sections of this rock, but recrystallization has been 
so complete that it probably destroyed all trace of algal cells, which 
are less than .01 mm. across. The forms may safely be attributed to 
probable algae. 

The top of the upper cherty beds is in nearly every respect a con- 
tinuation of the best magnetite bed, below the chert layer. It seems 
that the algal growth occurred at one stage in the accumulation of the 
conglomeratic ore. Possibly the accumulation never ceased, for the 
spaces between the algal "fingers" are filled with similar pebbly material. 

All this rock that has as much as 25 per cent magnetic iron can be 
concentrated to a product over 60 per cent iron, without grinding finer 
than about 100 mesh. The thin sections show a large proportion of 
the magnetite to be recrystallized grains of large size, and easily sepa- 
rated from the silica. 

Summarizing, it may be said that the upper cherty beds, with the 
exception of the algal chert near the top, constitute a unit in which 
the subdivisions are more or less arbitrary. The total thickness is 
between 120 and 160 feet with a little richer material near the top. 
The conglomerate is most conspicuous in R. 12 W. and is a distinguish- 
ing feature throughout the entire thickness of the upper cherty beds. 
While the thickness of these beds with relatively high magnetite con- 
tent apparently remains about constant throughout the entire Mesabi 
range, the conspicuously conglomeratic phases, containing within them 
the best concentrations of magnetite, may be thinner toward the west. 


The lowest of the upper slaty beds marks the end of most of the 
coarse conglomerate. There is no abrupt change, however. The gran- 
ule texture is seen in lenses in the slaty beds and there are even a few 
pebbles. Two features distinguish these beds from the preceding: 
first, the magnetite layers instead of being lenticular and compact, 
are straight thin layers of very fine grain, a set of thin magnetite beds 
alternating with the lenticular beds of granule texture; second, the 
coarser beds contain and may consist of nodules, or concretions dis- 
tinct from the matrix. Plate VI B shows these two features and indi- 
cates clearly the characteristic nodules. 

The development of nodules has probably played a more or 
less prominent part in the history of the formation. The nod- 
ules are not recognizable in drill cores, but they are very clear in 
the exposures and grade into the lenticular bedding, so as to suggest 

■ Walcott, Chas. D., Pre-Cambrian Algonkian algal flora: Smithsonian Miscellaneous Collec- 
tions, vol. 64. no. 2, p. Ill, 1914. 









that the very lenticular beds, even in the conglomerate, may have 
developed by concretionary action. It is noteworthy also that the 
nodules and matrix have the granules and even pebbles in them; so 
that the growth of a nodule apparently does not mean a replacement 
of all the material by a uniform precipitate from solution; rather it 
must have been more like a special cementation of a certain part of 
the beds. 

One of the most striking features noted in the outcrops of taconite 
on the east Mesabi is the septaria cracks in these beds. Septaria seem 
to have been described chiefly in carbonate concretions; and there may 
have been more or less carbonate in the iron formation — some still is 
to be found. The concretions now, however, are mainly quartz and 
magnetite, while the cracks are filled with white quartz. Plate VIII 
shows the appearance of the septaria. They are very useful in deter- 
mining the horizon of an outcrop. 

Several of the magnetite beds of this division have been crumpled 
or drag folded between more competent layers of chert, and more or 
less irregularly thickened (Plate XI A). 

There are abundant brown spots of garnet near the bottom of this 
bed, as compared with other parts of the magnetite rocks, but the 
garnet zone is too thin to be separately mapped. 

The abundance of thin magnetite beds makes the average of this 
bed worthy of consideration, though its concretions have only a small 
amount, and the finer grain requires finer grinding for good concen- 
tration. The magnetite bands, under the microscope, show swarms of 
minute quartz and amphibole inclusions and intergrowths, and Mr. 
E. W. Davis of the Mesabi Syndicate found that grinding must be car- 
ried to about 300 mesh to get a concentrate over 60 per cent iron. 

On the basis of the thin beds of fine grain, we have correlated these 
beds with Wolff's "upper slaty" beds, since he states that his terms are 
textural only and do not refer to composition. This report refers to 
them as "septaria beds." 

The beds above the septaria beds are similar in most respects, but 
septaria cracks have not been noted. They have less of the concre- 
tionary forms and granule structure, but the thin beds with a fair per 
cent of magnetite persist, growing a little leaner at the top. This upper 
part is not separated from the septaria bed in the general map. It has 
a relatively small number of magnetite layers, and grades into the 
beds above, (Aub 6 ). 

Taken together, the beds mapped as Aub 5 , are about 75 feet thick, 
except east of the Dunka River where they seem to be thinner. The 
outcrops show very little slaty character, except that the magnetite 
layers are very thin. Near Spring mine there is an outcrop at this 





horizon that weathers into thin slabs and breaks up much more readily 
than most of the taconite. 

The upper member of the formation lies on the slaty beds just de- 
scribed, with a rather sharp change in character. The magnetite is 
almost entirely lacking and the thin beds and granule texture have not 
been noted. Beds of white quartz and green silicates, from half an 
inch to 6 inches in thickness alternate to make a conspicuously banded 
light colored rock. Scattered grains of carbonate are more abundant 
than in the lower beds. There is little reason for describing this 
division as slaty, but it lies between slaty taconite and Virginia slate, 
and is not given a separate group name. It is everywhere about 50 
feet thick. 

The topmost bed of iron formation does not outcrop at many places 
and is too thin to show on the map. It is a calcareous bed about 10 
feet thick. Nearly every drill core that passes from Virginia slate to 
iron formation reaches such a bed just below the slate. This is not a 
pure limestone, but is more calcareous than any other beds of the iron 
formation. It is generally assumed that this marks the top of the iron 
formation. In contrast with the varying character of the iron forma- 
tion, the slate above is very uniform, so that this last change in 
material is a good horizon to select as the upper limit of the formation. 


The section given differs in some details from the sections given by 
Wolff, but there seem to be sufficient data for the correlation given in 
Figure 4. The persistent and characteristic "intermediate" slate is so 
well identified that it may be at once correlated with the slate in the 
eastern area. From this correlation it follows that the beds below the 
slate are much thinner at the east than on the main part of the range. 
Wolff shows the lower cherty beds about 280 feet thick; at Mesaba 
they are about 70 feet and in R. 12 W., about 30 feet thick. Accom- 
panying this change, and agreeing well with it, the Pokegama quart- 
zite disappears to the east. It is clear that active sedimentation began 
somewhat later in the east Mesabi than farther west. 

The persistence of the layer of calcium carbonate at the top of a 
series of four recognizable divisions of the iron formation is taken as a 
very strong indication that all the iron formation is represented in this 
eastern section. Van Hise and Leith 8 refer to the "known irregular 
alternation of iron-bearing formation and slate, both across and along 
the beds," as a cause of the varying width of the iron-bearing forma- 
tion. However, where the exposures are best and drill cores most 
numerous, there is no alternation that causes any doubt as to the loca- 

8 Op. eit., p. 174. 



tion of the limiting zone of carbonate. The reduction in thickness and 
width of outcrop is not explained by a lateral gradation of the upper 
beds into slate. Wolff, in calling attention to the variation in thick- 
ness in neighboring drill holes, suggests an erosion after the deposi- 
tion of the iron formation was completed, but the persistence of the 
upper beds makes this less likely than the variation in thickness of the 
lower beds. 

Thus the reduction in the total thickness of the iron formation from 
800 feet, as a large average, to 400 feet at the east, is largely attribut- 
able to the thinness or absence of the lower beds. The granite surface 
on which the iron formation was deposited, in R. 12 W., shows valleys 
about 80 feet deep. The upper beds, however, decrease somewhat in 

The value of such correlation has been emphasized by Wolff in his 
work on the hematite ores. 9 In summary it is found that the character 
of the hematite, whether blue and hard, or brownish-yellow and soft, 
is largely determined by the original horizon. Leith has noted a very 
constant relation also between the amount of phosphorus and the 
horizon; 10 the finer grained rocks, slate and paint rock, and the ores 
near them have high phosphorus. It is noted by both Leith and Wolff 
that the detail of differences in beds, and the correlation of beds, can 
be used to determine structures in the iron formation; and the slumping 
and folding so determined are guides to the occurrence of leached and 
enriched ore. 

Wolff observes further that the lower cherty and upper cherty hori- 
zons seem to be the most susceptible to enrichment into ores. Empha- 
sis should be placed on this fact here, for it is not an accidental or arbi- 
trary condition. The original taconite was richer in these horizons. 
It is evidently more likely that ore will be enriched to 55 per cent iron, 
if the original contained 30 per cent, than if it had less than 10 per cent 
iron. On the east Mesabi, interest centers in the quality of the orig- 
inal, unenriched taconite, and since the several beds are characteris- 
tically rich or lean this determination may guide the exploration 
farther west along the favorable horizons. 

Plate X shows that the several horizons are recognizable in drill 
cores and can thus be traced across the range. 

• Op. cit., p. 154. 

»• Leith, C. K.. The Mesabi iron-bearing district of Minnesota: U. S. Geol. Survey Mon. 43, 1903. 


B C. D. E. 




The Graham pits at Mesaba. — The two Graham open pits at Mesaba 
are near the main line of the Duluth and Iron Range Railway. They 
were until recently worked for hematite. The magnetite which may 
be assumed to have been the chief primary iron mineral is almost 
wholly oxidized in the workable ore. Very little attention is given in 
this paper to the process of enrichment of such ores. It is noteworthy, 
however, that the horizon of the ore bodies is shown to be that of the 
primary magnetite beds (Aub 4 and Aub 5 ). Septarian flint concretions 
occur near the tops of the pits. 

The Mayas mine. — The enriched hematite of the Mayas mine was 
apparently about worked out to the depth of the pit. Some ore is left 
in the bottom, as shown by recent drilling. The hematite concentra- 
tion at this point occurred near the horizon of the lower zone of mag- 
netite concentration below the intermediate slate. Recently the 
demand for manganese has led to renewed exploration of the pit and 
several layers with 6 to 30 per cent manganese have been uncovered and 
mined. These are to be attributed to secondary concentration. 

The Spring mine. — The ore of the Spring mine, like that at the 
Mayas mine, was concentrated at the horizon of the lower magnetite 
bed, below the slate. Oxidation has not been complete in all parts of 
the ore, as the remnants left in the pit contain considerable magnetite. 
The ore formerly mined, probably all of the best grade, may have been 
mostly oxidized. 

Hematite prospects. — Prospects and drill holes show hematite at 
several places east of the Spring mine, but all show a mixture of hema- 
tite with magnetite. None seem to have been rich enough to warrant 
development. Most of the prospect holes are in the lower magnetite 
zone, below the slate. An area in the N.W.}£ sec. 28, T. 59 N., R. 14 W., 
received special attention. 

The enrichment of hematite ores. — The leaching, oxidation, and 
enrichment of these ores was no doubt accomplished by the processes 
affecting the ores farther west, and studied in detail by Leith 1 and 
Wolff, 2 both of whom recognize a relation of the enrichment to folds 
and jointing. The eastern hematite deposits have not been examined 
in detail, but it is noteworthy that the Spring mine, the most easterly 
mine worked, is in a rather sharp fold. It may also be noted that the 
lower ore horizon is near the granite, and likely to receive alkaline 

1 Leith, C. K., The Mesabi iron-bearing district of Minnesota: U. S. Geol. Survey Mon. 43, 1903. 
* Wolff, J. F., op. cit. 



solutions that would leach out silica; while the upper ore horizon along 
much of the belt is nearer to a swampy tract where the acid water 
would dissolve out iron rather than leave the ore enriched in iron. 

It is commonly supposed that the enrichment of the soft hematites 
of the Mesabi occurs without much effect on the state of oxidation of 
the iron. However, it is clear that at these eastern mines the iron of 
the fresh taconite was mostly in magnetic form. Outcrops on both 
sides of the hematite body contain magnetite. Since this magnetite 
has been altered to the ferric iron producing the secondary hematite 
ore bodies, it is possible that magnetite may have been the chief pri- 
mary oxide all along the range. This is indicated by some preliminary 
tests on drill cores from the central and western parts of the range. 



The minerals of the deposits. — Practically all the minerals men- 
tioned as occurring in the iron formation, are found in the magnetite 
bodies. Probably over 95 per cent of the beds Aub 4 and Aub 5 consists 
of amphibole, quartz, and magnetite. To the west, some hematite is 
intergrown with the magnetic iron oxide. 

The magnetite is rarely pure, but the impurities are the intergrown 
quartz and amphibole. There are very small amounts of pyrite, and 
analyses show that the phosphorus is low in the beds containing the 
most magnetite. Titanium has not been found even in the deposits 
close to the Duluth gabbro, which contains much titaniferous magnetite. 

Chemical composition of the magnetite rock. — Van Hise and Leith 1 give 
the following analysis of the average rock of the east Mesabi (apparently 
the samples are actually from the Gunflint Lake region) which is tabu- 
lated, for comparison, beside the average taconite of the west Mesabi. 

Average Composition of Amphibole- Magnetite Rock and Taconite of the Mesabi District 

Amphibole-Magnetite Taconite 

Si0 2 60.51 58.71 

A1 2 3 1.20 .54 

Fe 25.22 25.71 

MgO 52 

CaO 67 

H 2 Small (Ignition) 1.96 

P2O5 05 (P) .021 

S 59 

Mn0 2 92 

TiQ 2 None 

As compared with these analyses, which represent the average of 
the whole formation, the following results were obtained by J. H. 
McCarthy, at the Minnesota School of Mines experiment station, on 
an average sample taken from the upper cherty beds, where magnetite 
is more abundant than in the average of the whole formation. 

Analysis of an Average Sample of Cherty Beds of Magnetite Rock near Sulphur 

Si0 2 45.17 

AI2O3 ' 0.67 

Total Fe 33.80 

CaO 3.02 

MgO 2.39 

C0 2 0.58 

Ti0 2 Trace 

Mn 0.61 

Soluble Fe 29.20 

P 0.049 

1 Van Hise, C. R. ( and Leith, C. K., The geology of the Lake Superior region: U. S. Geol. Survey 
Mon. 52, pp. 181, 185, 204. 



Textures. — The textures of the magnetite and its matrix are prob- 
ably indicative of their origin. The most conspicuous features are the 
abundance of conglomerate in the upper cherty beds and the concre- 
tionary beds above the conglomerate. (Plates V B and VI B.) 

The granule texture of the taconite in layers alternating with mag- 
netite has been studied in detail by Leith, who ascribes most of it to 
the occurrence of greenalite in the fresh original rocks. In the altered 
rocks of the main eastern area, greenalite is entirely recrystallized. 
Some granules of amphibole may represent original greenalite, but 
there are many rocks in which the texture gives an impression of a 
series of fragmental grains of varying size. The small grains are less 
than .05 inch, but they seem to have similar characters in all sizes up 
to 6 inches. Most of the grains are magnetite in large part, though the 
magnetite may be intergrown with or intimately related to the quartz 
and amphibole. Many of the pebbles are cherts or other siliceous 
material. Many of the siliceous pebbles have a richer magnetite rim 
or border. A few are zoned in alternating bands of magnetite and sili- 
ceous matter, as if altered by diffusion. Many are fractured in an 
irregular network of cracks resembling septaria, with white quartz 
filling. These remarks apply to the microscopic grains as well as to 
the larger pebbles. (See Plate IX.) Some of the larger pebbles con- 
sist of conglomerate made of granules and smaller pebbles. 

Some of the black pebbles show a dull appearance resembling gra- 
phitic rather than ferruginous minerals. Many of these have adjacent 
to them in the granular matrix a zone of lighter color than the average 
matrix. Some of the pebbles, themselves composed of granular mate- 
rial, are hardly distinguished from their granular matrix. 

The magnetite layers alternating with the conglomerate and gran- 
ular taconite are compact and contain more or less quartz and amphi- 
bole intimately mixed. Many of them, however, seem fairly pure. 

The concretions, bearing septaria of white quartz, have been de- 
scribed above as guides to horizons near the center of the magnetite 

All these textures tend to become obliterated by recrystallization, 
northeast of Dunka River. The conglomerate, however, is recog- 
nizable almost to Birch Lake, while the chert becomes an aggregate 
of quartz with grains up to half an inch in diameter. The coarser grain 
has an economic significance, because in the magnetic separation the 
grains of impurities must be separated by the grinding, and the coarser 
ores require less grinding to give a clean concentrate. It is probable 
also that the reduction in porosity during crystallization, from 5 per 



cent to less than 1 per cent, 2 is responsible for the lack of much weath- 
ering in the eastern areas. 

Stratigraphy of the magnetite. — It has been found that the amounts 
of magnetite in the several beds of the iron formation here dis- 
tinguished are fairly constant over a wide range of territory. The pre- 
liminary estimates made on field observations have been checked by a 
number of assays, and notwithstanding considerable local variation in 
any given bed, the concentrations of iron prove to be at fairly uniform 
horizons. Figure 3 gives a generalized estimate of the horizons of 
magnetite concentration. It may be assumed at once that any large 
area of the conglomerate bed and probably the lower member of the 
upper slaty beds will bear about -20 per cent of iron in magnetically 
separable form wherever they are located east of Mesaba, except in the 
rare case of weathering to hematite. In detail, however, these beds in 
large areas probably vary from 12 to 28 per cent in magnetic iron. 
Drilling has already shown some variations, and there are concealed 
areas as large as a square mile. Furthermore it may be assumed that 
the smaller beds just above and below the cherty beds with algal struc- 
ture, will be considerably richer than 20 per cent iron in magnetite. In 
the magnetite beds, the iron in minerals other than magnetite increases 
the total iron content to about 35 per cent, but this is not considered 
available, so long as the recovery of iron depends on magnetic con- 
centration. The smaller magnetite layer below the intermediate slate 
is perhaps less uniform, and certainly not exposed in so many places, 
but it is similar in grade to the thicker upper zone. 

The importance of the detailed stratigraphy reported in Chapter 
IV becomes apparent as soon as the deposits of magnetite of this grade 
prove to have commercial value. Outcrops along the range are about 
as numerous in one horizon as another, and the identification of any 
one horizon will be a definite guide to the location of the magnetite 
deposits. The persistence of certain features such as the algal struc- 
ture and septaria cracks, or even the intruded diabase sills, makes it 
possible to determine the limits of the magnetite zone with unusual 
accuracy even without drilling. A small amount of drilling confirms 
the field estimates in most cases. The drill cores furnish still other 
horizon markers. Mr. Fred Jordan, one of the engineers conducting 
the explorations near Sulphur, has found that correlation is possible on 
the basis of the per cent of soluble iron in the tailings from magnetic 
concentration. This iron may be in the form of fayalite or siderite, or 
possibly hematite. The fayalite was most often noted in the field. 

* Van Hise, C. R., and Leith, C. K., op. cit., p. 554. 



It is to be expected from the greater thickness of the iron formation 
farther west, that there will be some variation in the thickness of mag- 
netite zones. The irregularities of the bedding, on a small scale, might 
suggest that the deposit would be erratic. Nevertheless it may again 
be emphasized that in spite of local variation the average magnetite 
content of certain beds is high enough to be of interest. The thickness 
of the main zone carrying conglomerate or septaria is 200 feet, and as 
this increases to the west, there may be included lean beds. A certain 
amount of drilling is necessary to determine the favorable locations. 
Outcrops and drill records, however, in R. 12 W., show a conglomerate 
zone 100 feet thick near the center of the iron formation. Wolff shows 
a similar section with conglomerates through a zone of 100 feet in R. 
18 W., and 95 feet in R. 20 W., 35 miles west of the eastern Mesabi 
area. These records make correlation fairly safe. 

Distribution of the magnetite deposits. — The magnetites east of 
Mesaba were known before the deposits of the Mesabi range west of 
Mesaba were developed. The changes from hematite ore bodies under 
drift, farther west, to harder magnetite in good exposures farther east, 
are conspicuous. Certainly the proportion of hematite in the formation 
grows rapidly lower to the east of Mesaba. Nevertheless when atten- 
tion is directed to the leaner, unaltered taconite containing magnet- 
ite, and the secondary concentrations are ignored, the amount of fer- 
ruginous material that can be concentrated magnetically from fresh 
taconite is only slightly less near Mesaba than farther east. West of 
Mesaba fresh taconite, mostly from drill cores, is available from 
Aurora, Gilbert, Virginia, Eveleth, Hibbing, and Coleraine. Con- 
centration tests on these samples indicate about as much magnetic 
iron in these cores as there is in the area between Mesaba and Birch 
Lake. The data at hand indicate that the horizons of primary iron 
concentration (now magnetic) are probably just as thick as they are 
east of Mesaba. The western areas are more interrupted by secondary 
leached zones of hematite, and are much more deeply covered with 
drift, but there can be no doubt that large bodies of lean taconite in 
the western areas carry over 20 per cent iron in form for magnetic con- 
centration. Furthermore, there are in the western, as in the eastern 
area, two main horizons, one above and one below the "intermediate 
slate." Through the central part of the range each of these magnetite 
zones is over 100 feet thick. 

As an illustration of the masses which have been drilled without 
recognition of magnetite, Figure 5 is presented. It is sketched from a 
reexamination of cores from drilling, in which a fair body of hematite 
was discovered — perhaps two or three million tons. Two magnetite 



bodies are outlined, one about two million and the other ten million 
tons. These are an average of 55 feet thick, under a drift cover of 65 
feet, — less favorably situated than the deposits east of Mesaba, but 
still worthy of note. The samples of these cores were ground to 100 
mesh and concentrated magnetically, 2.83 parts of ore giving 1 part 
concentrates with an iron content of over 64 per cent. 




This discovery that magnetite is a prominent constituent in the 
fresh taconite of the west Mesabi was unexpected, because of previous 
reports of the dominance of hematite west of Mesaba. 3 It was also sup- 
posed that magnetite is relatively stable under leaching and would 
not ordinarily alter to a soft hematite. It is very evident, however, that 
ferric oxide has formed east of Mesaba, by alteration of magnetite. 
The occurrences of abundant magnetite in the west part of the range 
in the fresh unweathered iron formation suggest that the protore there 
also contained more magnetite than hematite. 

Magnetic mapping. — The persistence of magnetite across the range 
and its localization in certain beds of the iron formation might indicate 
the probability of a successful magnetic search for these horizons. 
Leith, however, recorded observations along the range indicating much 
more magnetic effect in the east Mesabi than on the main range and 

* Magnetite has recently been noted by J. F. Wolff, Recent geologic developments on the Mesabi 
iron range: Lake Superior Min. Inst., Proc, vol. 21, pp. 229-57, 1917. 



concluded that there is much more magnetite east than west; he also 
noted that no special horizons were detected by the needle unkss it 
was the zone just above the Pokegama. He found high readings in 
places near the contact of hematite and fresh taconite. 

Detailed dip needle and dial compass readings have been taken on 
lines one quarter of a mile apart, all along the east end of the range, 
and while the deflections show the boundaries of the iron formation as 
a" whole, it has not been found possible to relate them very definitely 
to the several horizons. No line of readings crossed the iron forma- 
tion without finding some high dip needle deflections, and the dial 
compass was affected somewhat erratically, mostly in the belt of the 
better magnetite deposits, but it has not been possible to make much 
use of the data in the detailed mapping or in the search for the richest 
magnetite. (See the description of the Iron Lake area, pages 52 and 53). 

In explanation of the rather weak attraction which is shown by the 
iron formation of the main part of the range for the magnetic needles 
used in exploratory work, the following points are mentioned : 

1. It has been observed that tabular bodies of magnetite with a 
low dip to the south, such as is shown by the Biwabik iron formation, 
may cause very little disturbance of the earth's magnetic field, and 
hence have little effect on the magnetic needles. 4 

2. The oxidation of the magnetite to ferric oxides to a depth of several 
hundred feet in many places makes the attraction at the surface weaker. 

3. The burial of the iron formation beneath a great thickness of 
glacial drift has the same effect. 

Method of estimating reserves. — With a little more drilling it seems 
likely that the continuity of the zones of iron concentration through 
the range may be very definitely established. If it is agreed that the 
primary magnetite concentrations occur as strata, drill holes spaced 
several hundred yards apart will give a good check on the location of 
ore and the amount of drift cover. The estimation of reserves becomes 
a problem of stratigraphy rather than one of drilling. Detailed drill- 
ing will be entirely unnecessary unless it is found to be the best method 
for the study of the thickness of drift, and two or three variable fea- 
tures which interrupt the formation locally, viz., dikes, faults, and 
secondary alteration to hematite. With due allowance for these fac- 
tors, the volume of ore can be calculated with very little drilling. 

The quality of the ore also may be estimated in the case of such 
primary beds with much less drilling than in the case of enriched ores. 
While there are variations in the quality of beds, the changes are much 
less abrupt than in the case of enriched ores, and a few holes should give 
a fair indication of the average ore. 

* Broderick, T. M., Some features of magnetic surveys of the magnetite deposits of the Duluth 
gabbro: Econ. Geology, vol. 13, no. 1, pp. 35-49, 1918. 






Commercial situation. — It is, of course, a commercial matter to 
decide whether or not any particular deposit of magnetite is to be 
classed as ore. Up to this time reports on the Mesabi magnetite have 
been unfavorable; but the economic features of these deposits indicate 
that as the reserves of richer hematite ores are depleted they will be 
worthy of more and more consideration. 

The commercial situation naturally varies from day to day. The 
world war started an increase in the demand for iron, and the statis- 
tics issued by the United States Geological Survey show that the pro- 
duction in the Lake Superior district increased greatly in 1915 and 1916. 

Iron Ore Produced in Lake Superior District, in Gross Tons 

Prices have advanced as the demand continued, and in spite of labor 
and transportation difficulties new reserves have been discovered and 
new mines opened. Some old abandoned mines have been reopened, 
and it has been said that if an iron mine can not be made to pay under 
such favorable circumstances there seems to be little hope for its imme- 
diate future. 1 It seems likely that a well equipped plant might have 
made a profit under these conditions, if it had been already established 
on the east Mesabi, but the time was not as favorable for building new 
plants as for stimulating production from those already built. Two 
small quarries and the experimental mill at Duluth have been operat- 
ing recently. 

Per cent of iron and volume of material. — The grade and amount of 
ore must be considered together. Specimens a few inches in diameter 
may show over 70 per cent iron, but no large volume of such ore is to 
be expected. Van Hise and Leith 2 dismiss the magnetites with the 
statement that they are a few inches to a few feet thick in layers with 
other rocks less rich in magnetite. It is true that beds with 30 per cent 
or more of iron in a form for magnetic concentration are relatively 
thin, but the beds with 20 p^r cent of iron (including some smaller 
beds with over 30 per cent) make up an immense tonnage. The ton- 
nage and quality of material necessary to constitute workable ore 
must be determined by mining and milling costs and prices of ore. 

1 Burchard, E. F., Our mineral supplies, irou- U. S. Geol. Survey Bull. 666, V, p. 3. 
* Van Hise, C. R., and Leith, C. K., Geology of the Lake Superior region: U. S. Geol. Survey 
Mon. 52, p. 185. 






This report gives simply an indication of the general situation as to 
grades and quantities. 

The upper eherty beds, Aub 4 , contain near the top two layers of 
ore with about 30 per cent magnetic iron, and even more in some places. 
The lower of these is in places 20 to 30 feet thick and the upper about 
10 feet thick. They are separated by a few feet of leaner cherty or 
jaspery material with a conspicuous algal structure, so that they are 
easily recognized. If this leaner bed is quarried with the rest, to get a 
working face 40 to 50 feet high, the grade might be reduced to a little 
below 30 per cent. The lean bed is easily distinguished, however, and 
could be cobbed out. The volume of such a bed several miles long 
and 50 feet thick must be very great, but since it dips at an angle of 
about 5° there is a relatively narrow belt in which the ore can be 
worked, with a high quarry face, without stripping orf the overlying 

The beds mapped as Aub 4 and Aub 5 , taken all together with rich 
and lean beds, and rich and lean areas, may contain as much as 20 
per cent of iron available for magnetic concentration. Lean beds, 
especially near the top, will be partly balanced by the richer beds near 
the algal chert. Since the drilling shows some areas much richer than 
others along the belt, it must be remembered that drilling may be 
needed to select a favorable area. The outcrops shown on the map 
extend along a belt about 20 miles long and the average width is about 
2,000 feet. With a dip of 5° it may be assumed that the northern edge 
is too thin to be considered for a distance of 400 to 500 feet from the 
border. South of that, the magnetite beds cover the area mapped to 
a depth of from 50 to 200 feet, with only a small amount of drift cover 
at most places. A rough calculation indicates that the east Mesabi 
range from Birch Lake to Mesaba contains about 1,500,000,000 tons 
of this magnetite formation within 100 feet of the present rock surface, 
without any bed rock cover. If such a reserve is to be considered as 
"available at present," it will more than double the present reserves of the 
range. The reserves of similar grade farther west are less accessible 
since they are covered more deeply with drift, but they are very large. 

Van Hise and Leith 3 estimated in 1911 that the total present avail- 
able reserves on the Mesabi were nearly 1,500,000,000 besides the 
"wash ores." If the lean bodies of "wash ore" of the west Mesabi 
are to be included, these magnetic ores of the east Mesabi may well 
be added also. They have certain advantages over other lean ores. 
For example, Van Hise and Leith 4 include in the reserves available 
for the distant future, ores that have 35 to 40 per cent iron, regardless 

s Op. cit., p. 489. 
*Op. cit., p. 491. 



of whether or not they are concentrating ores, but add that very little 
of this ore will be used for a long time, except by mixing with high- 
grade ores. The advantage of the magnetites lies in the fact that they 
can be concentrated to a very high grade. This gives a supply of high- 
grade concentrate which will make it possible to continue mixing in 
the low-grade ores, long after the naturally high-grade ore is exhausted. 
Much of the magnetite bed here considered now contains about 35 
to 40 per cent total iron, but only part of this iron occurs in the min- 
eral magnetite. 

Topographic situation. — Most of the outcrops of magnetite are on 
the south slope of the Giants Range and extend with minor undulations 
down the slope, descending about 100 feet in half a mile. Some of the 
exposures on the lower southern side are near the level of the swamps 
or valleys in the slate and not favorably situated for mining, but a 
large part of the formation stands at high levels and can be mined in 
open cuts with surface drainage. 

Structures. — The beds dip gently and it will not be very difficult to 
follow the valuable beds down the dip. The bedding planes, though 
very irregular, furnish surfaces of easy parting. 

The joint system in the conglomerate beds is especially favorable 
to shattering the rock with heavy blasts "against resistance" of pre- 
viously broken material. The joints are nearly vertical and run prom- 
inently about north and south in most places, spaced 1 to 6 inches 
apart. They are crossed by another set about at right angles. Other 
sets of joints have been locally noted. Plate XI B shows the promi- 
nence of the joints where weathered, but traces of the same structure 
make the rock break well even at considerable depth. 

Magnetism. — The fact that the ores are of magnetic oxides has led 
to experiments with magnetic milling methods which are so promising 
that ores of much lower grade than heretofore become of interest. 
Throughout the east Mesabi the beds designated as magnetite deposits 
have their iron largely in a form that can be magnetically concen- 
trated. Nevertheless the ore of some of the deposits shows under the 
hammer a red powder and contains some hematite. 

The chemical tests of the several divisions of the iron formation 
have not been carried far enough to determine why some beds consist 
largely of iron silicate — fayalite — while others with similar amounts of 
iron consist of magnetite and quartz. The fayalite is most commonly 
developed in a quart z-amphibole matrix, but some of the samples have 
magnetite also, and may be as rich as the rest of the deposit. The 
matter of combination may be determined by the state of oxidation or 
by the presence of other elements. Whatever it is, the results are clear; 



some beds have considerable iron in magnetic form, and these are the 
beds distinguished in mapping. 

Size of grain. — The grain varies in different beds as well as in dif- 
ferent areas, with more or less relation to the distance from the gabbro. 
The coarser grain is favorable, requiring less fine grinding for the mag- 
netic separation of a rich concentrate. The finer grains remain attached 
to quartz and silicates which then get into the concentrate and reduce 
its grade. 

The coarse-grained beds, favorable to concentration without the 
expense of grinding finer than 100 mesh, are the conglomerate beds, 
Aub 4 , and all those that have magnetite east of Dunka Rivei. 

The impurities that may be retained in the concentrates, if not 
ground fine, consist not only of silica, but of phosphorus, which is even 
more deleterious. The concentrates normally bring a premium because 
of their low phosphorus. Apparently some parts of the deposits can 
be made to yield, without special treatment, concentrates so low in 
phosphorus that the premium is high. Sulphur is normally removed 
without extra care. 

Hardness. — The t aconite is one of the toughest, hardest rocks that 
is quarried. It will always be hard to drill, but the jointing is a great 
aid in breaking up the rock after blasting. 

Milling. — The Mesabi Syndicate have had the benefit of the work 
of men widely experienced in magnetic concentration, and have built 
a mill at Duluth in which they have conducted tests of their ore as 
well as several new devices for milling, and various materials for sin- 
tering. The mill has been operating to capacity, producing low-phos- 
phorus concentrates. 

Grade of product. — The fineness of grinding and other variable 
processes in milling permit the operators to determine in advance about 
what quality of concentrate will be produced. By manipulation of the 
process, then, the mill can produce any grade of concentrate for which 
there may be a demand, even those which draw a premium for unu- 
sually low phosphorus, and high iron, contents. 

In the fall of 1918, a trial cargo of low-phosphorus sinter made from 
east Mesabi ore was shipped from Duluth to the Midvale Steel and 
Ordnance Company's blast furnace at Coatesville, Pa. The cargo 
analysis shows an iron content of 63 per cent, with .008 per cent of 
phosphorus. It is expected that the concentrates made from these 
ores to supply the normal demand will run over 60 per cent iron and 
from .020 to .025 per cent phosphorus. 

Transportation. — The distance from market is about equal to that 
of other Mesabi ores. The Duluth and Iron Range Railroad now has 
a temporary line along most of the magnetite belt. 



Cost of exploration. — It has already been suggested that when 
enough drilling has been done to satisfy the explorers that the ore beds 
are continuous, a very few drill holes will give all the data needed as 
to the quality. It should be much cheaper to explore a property of 
this regular bedded type, than the irregularly leached deposits of soft 
ore of the Mesabi. 

Scale of operations. — There is plenty of magnetite so that mining 
operations can be undertaken on as large a scale as at any mine in the 

Suggestions for conservation. — 1. Sampling. Since 1891 when the 
rush came to the rich ores of the Mesabi range, it has become a fairly 
established custom to sample drill cores for analysis wherever there was 
an appearance of enrichment, even if the ore seemed to be lean. On 
the other hand, hard unaltered looking taconite was of no interest and 
any lump of core might be thrown in a box as a fair sample for the 5 
or 10 feet of core from which it came. In the magnetite bodies all the 
core looks fresh and unaltered. Nevertheless it is important to have a 
fair sample. Many of the old cores which have been examined in this 
work indicate large bodies of magnetic material, but the accuracy of 
the samples is very uncertain. 

2. Magnetic tests. Before a mining property is wholly abandoned 
for lack of ore, the cores with magnetite should be tested by magnetic 
concentration to see if any very large body of 20 to 30 per cent mag- 
netite is in favorable position for development. 

3. Location of dumps. It has often happened on the Mesabi range 
that the dump piles from stripping operations and the waste of mining 
were placed where they had to be moved later, to get some good ore 
that was unknown at the time. We suggest therefore, that even now 
before the magnetites are worked, dumps should not cover the better 
magnetite bodies. It is uncertain how long it may be before the mag- 
netite will be wanted, but when the time comes, the ore should not be 
inaccessible because of our carelessness in locating dumps. (See Fig- 
ure 5, illustrating a neglected magnetite body which might be care- 
lessly buried under so much waste that it would not pay to remove it, 
to get the magnetite.) 



Introduction. — The bedded and banded character of the iron forma- 
tion taken in conjunction with its mineral composition indicates clearly 
the sedimentary origin of the material. 

Agent of transportation. — Cherts and the several primary sediments 
of iron, on the Mesabi, are materials known to be precipitated from 
water solution, and it is generally agreed that in those parts of the 
formation where clastic grains are lacking, it may be assumed that the 
iron-bearing sediments were deposited from water solution. No 
mechanical sediment is known which closely resembles them. The 
solution of silica may have been facilitated by the presence of alkalies; 
that of iron would be more likely in the presence of acids. If carbonate 
minerals were more abundant (as in some other ranges) an alkaline 
bi-carbonate solution might be suggested as the most probable com- 
bined solvent. This kind of solution is known to have had an igneous 
origin in some places. Incidentally some alkali may have been derived 
from the granite near by. 

Source of the iron. — The source of the iron in the water solutions is 
naturally a matter of great interest. This question is discussed at 
length by Van Hise and Leith. 1 They conclude that for the iron of 
unique, thick, extensive iron formations" like this, it is necessary to 
appeal to contemporaneous basic igneous rocks. Some of the iron may 
have been derived from the action of water on the rock, either hot or 
cold, but large parts of it may have been carried in solution by direct 
magmatic emanations. There is less contemporaneous igneous rock 
near the Mesabi than in most iron districts, but there was a large 
amount of igneous activity from the Cuyuna range in Minnesota, to 
the Marquette district in Michigan, and probably elsewhere. 

Deposition. — L Precipitation. — From any assumed solution, it is 
easy to suggest reactions that will precipitate silica and iron. 2 The 
direct evidences of the actual form of the precipitate and the nature of 
the precipitating agent are very slight. There are the supposed algal 
forms in the cherts and jaspers indicating organic action. There are 
also the graphitic cherts and slates indicating organic material. 
Steiger 3 reports organic matter in some greenalite rock. Since it seems 
that organisms could live at that time and organisms are known to 
precipitate iron at the present time, it may be assumed that they 

1 Van Hise, C. R., and Leith, C. K., The geology of the Lake Superior region: U. S. Geol. Survey 
Mon. 52, pp. 506-18. 

2 Van Hise, C. R., and Leith, C. K., op. ciL, pp. 519-27. 

8 Steiger, George, See analysis in U. S. Geol. Survey Mon. 52, p. 167. 



played a part also in the deposition of the iron of this formation. This 
is not certain, because the deposits seem to have been more or less 
modified since their first deposition, and their original form is in doubt. 
Leith has shown that material of similar composition and character 
can be produced by inorganic action in the laboratory, so that the 
action of organisms is not essential to precipitation. Current studies 
however seem to indicate that if a reaction may occur both organi- 
cally and chemically, it occurs more rapidly and completely by the 
action of organisms. It is therefore believed that organisms pre- 
cipitated not only cherts, but at times, highly ferruginous cherts. 
Harder and Johnston refer to the possible precipitation of ferric oxide 
and silica, as the primary deposit s - of the iron formations, associated 
with more or less organic matter. 4 The iron of the Biwabik formation 
was probably in a combination of a ferrous silicate and carbonate, and 
a ferric oxide, in proportions varying according to varying conditions. 

2. Conditions of deposition. The extent of the magnetite deposits 
and the associated sediments indicates deposition on a broad sea bottom. 
The beds lie above a sand and below a clay, as a part of a series of sedi- 
ments, probably formed in shallow water. Van Hise and Leith 5 con- 
sider the water probably shallow, because of the predominance of sili- 
cates over carbonates, the structural differences from the Cuyuna iron 
formation, and the lack of outcrops farther north. An additional evi- 
dence appears in the flat form of pebbles in the central conglomerate. 

3. Texture of the deposit. The cherts with algal structure appear 
to be primary in their form and texture. Much of the remainder of 
the formation is filled with granules, pebbles, concretions, and nodules, 
indicating more or less reworking since precipitation. Throughout 
most of the iron formation the granules uniformly show certain min- 
eral and structural peculiarities which have led to the suggestion that 
they are not ordinary fragmental sedimentary grains. (See Plate 
IX D.) Leith has argued that the granules of greenalite are forms that 
may develop in colloidal precipitates by surface tension; or, as he sug- 
gested in his earlier paper, by the replacement and coating of organic 
remains, in some such way as the granules of the Clinton iron forma- 
tion are supposed to have formed. No doubt there are granules that 
are best explained as forms resulting from surface tension on precipi- 
tates, but there are also many pebbles which could not have formed 
from any process other than mechanical wear. Associated with frag- 
mental pebbles, granules of similar fragmental origin are certain to be 
numerous. The writers believe, therefore, that much of the texture 

* Harder, E. C, and Johnston, A. W., Geolo'gy and iron ores of the Cuyuna district, 
Minnesota: U. S. Geol. Survey Bull. 660-A, p. 16, 1917. 

s Van Hise, C. R., and Leith, C. K., op. cit., pp. 214, 604, and 613. 



has been modified since precipitation. Over half of the main magnet- 
ite deposit of the east Mesabi has the characteristic pebbly texture. 

4. The alternation of deposition. The alternation of sediments of 
several sorts in the iron formation as a whole indicates a probable alter- 
nation in the source of supply. The larger alternation of slate and 
chert shows that deposition occurred under different conditions. On 
the other hand, the smaller, more often repeated alternation of mag- 
netite with chert having a granule or pebble structure is not taken to 
mean any essential difference in the source of supply or agents of trans- 
portation, but rather an alternation of conditions acting on the material 
already deposited. 

Primary modification of the deposit. — As has been suggested, the 
repeated alternation of material in the beds now found is believed to be 
a significant fact in connection with the history of the iron formation. 
There are hundreds of alternations of fine magnetite and coarser frag- 
ment al layers. Rhythmic sedimentation is in some cases due to a 
rhythmic supply of differing materials. In a broad way, the supply 
may be ferruginous at one time, slaty at another, and cherty at 
another. If the material had been derived from volcanic sources, as 
has been suggested, it seems improbable that these supplies would 
alternate as many times as the sediments indicate. It is unlikely that 
there were so many successive flows. Volcanic rhythms should pro- 
duce alternations on a large scale. Furthermore, climatic rhythms are 
also larger features. The detailed alternation of beds from a tenth of 
an inch to 6 inches thick is more likely attributable to seasonal or other 
occasional changes in conditions. These changes would affect a chem- 
ically depositing sediment only if in shallow water, and emphasize the 
conclusion reached above as to the conditions of deposition. 

The conglomerate and granular beds of the lower parts of the mag- 
netite zones are very suggestive as to the conditions that must have 
prevailed. Most of the pebbles are flat, and vary widely in the degree 
of rounding. They include no fragments of the neighboring granite 
and schist, but many closely resembling the underlying beds of iron 
formation. Their size and angularity indicate no distant source. The 
irregularity suggests washed lumps of partly hardened mud, or pre- 
cipitate. The most probable source of the pebbles is a part of the iron 
formation itself. However, no signs of erosional unconformities have 
been found in the iron formation. Such a conglomerate is to be classed 
as intra) 'or motional, — one formed of material recently deposited and 
without any extensive transportation. Such conglomerates are 
reported from several limestone formations with characteristics that 



are so similar as to warrant giving this siliceous bed the same name. 6 
The conglomerate part of the formation, as well as some of the gran- 
ules, may therefore be attributed to a recent precipitate more or less 
broken up by wave action. The freshly broken fragments would be 
large and angular pebbles, those washed about for a longer time would 
be more rounded, and the smaller grains would become rounded granules. 

Most of the pebbles are magnetite. This might mean that an orig- 
inal ferruginous layer resisted the breaking action of the waves longer 
than some of the freshly deposited silica. Other pebbles, however, 
suggest another explanation. Many pebbles are cherty with a border 
of magnetite. It seems that at the time the pebbles were rounded (or 
possibly since then) some process enriched them in iron. Since it is 
believed that the pebbles formed in shallow water, the logical expla- 
nation is that standing water acted on the ferruginous chert, enriching 
it, much as the ores of the Lake Superior region are being enriched at 
the present day, — by solution of silica, and possibly, but only to a 
small extent, by deposition of iron in its place. 

This assumption of leaching of silica in a sea where silica was 
accumulating does not appear plausible unless, as was true in this 
case, some alternation of conditions is indicated. In a shallow sea the 
contribution of iron and silica from a magmatic or other source may 
furnish plenty of material for deposition. But at a distance from the 
source any addition of fresh water, say from a heavy rain on the ad- 
joining land, would crowd back the depositing solutions and replace them 
by water that would dissolve and oxidize the deposits, until diffusion and 
convection again brought in the stronger solution. 7 The volume of fresh 
water which leached the silica may not have been great enough to modify 
the composition of the depositing solution very much, after it was 
mixed by diffusion and general circulation. So the next deposit of 
iron is not necessarily leaner, or more siliceous than the first. Occa- 
sional storms might agitate the waters enough to break up the deposits 
and round the grains, but the special richness of the conglomerate in 
iron oxide is thought to be due to a primary leaching of the silica. The 
enrichment is considered primary because it is believed to have 
occurred before the overlying layer was deposited. While this enrich- 
ment is chemically the same, it is not such an enrichment as has 
occurred on the main Mesabi range in late geologic periods. That 
enrichment has been accomplished by circulating ground waters, and 
may be as deep as the waters carry solvent action. The primary enrich- 

« See for example, Foerste, A. G., Intraformational pebbles in the Richmond group: Jour. Geol- 
ogy, vol. 25, p. 289. 

* Van Hise and Leith note that some greenalite was probably oxidized at the time of deposition. 
Op. cit., p. 537. 



ment here considered probably occurred at the bottom of shallow 
standing water and its effects were superficial. Nevertheless if each 
thin bed is superficially enriched, the formation as a whole shows the 
effect. From this point of view the magnetite pebbles may be those 
which were most thoroughly leached. There may be an analogy in the 
Cretaceous hematite conglomerates of the west Mesabi, but this is 
doubtful. The best magnetite is so characteristically associated with 
conglomerate that there is a strong suggestion that the taconite has 
rich layers only where leached. 

In continuation of this argument, it is well to consider the end 
products of such processes of leaching and wave action as are here 
suggested. Eventually the silica might all be leached from a surface 
layer, and the residual iron compounds, thoroughly oxidized, might 
be pretty well pulverized. The conditions in deep water may also be 
such that a little leaching of silica might occur, leaving only a layer of 
iron oxide along the bottom. When deposition was resumed, this 
powdery ferric oxide, however formed, would lie on the irregular bottom 
as a layer of relatively pure ore. This is exactly the condition of the 
magnetite beds now alternating with conglomerate. 

The suggestion here given as to the origin of the pure magnetite in 
no way conflicts with the possibility, suggested by Van Hise and Leith, 8 
that some oxides may have been precipitated directly in very pure 
form. Precipitated oxides, however, would be expected to alternate 
with precipitated chert rather than with a conglomerate. 

After any accumulation of oxides, the next precipitate would tend 
to cement any .such layers into fairly firm masses, and the next storm 
would break them up into pebbles of iron oxide. This is the effect sug- 
gested by such outcrops as are shown in Plate IV A, where a bed of 
magnetite stops abruptly and there are near it some fragments that 
look as if they were just broken off. Thus both the enrichment of 
chert pebbles and the breaking up of greatly enriched iron oxide beds 
furnish ore pebbles to the conglomerate. 

The development of nodules in the magnetite deposits is probably 
not a phase of primary deposition, but is believed to be an effect that 
developed soon after deposition. In the main conglomerate beds no 
clear nodules have been noted, but in the thin beds above the algal 
cherts, the structures can hardly be given any other interpretation. 
(See Plate VI B.) 

Both silica and iron are known to be subject to concretionary re- 
arrangement, and on the west Mesabi some common chert nodules may 
be seen. There are a few small concretions containing both iron oxide 

• Op. ciL, p. 527. 



and quartz in concentric bands. It is not generally supposed, how- 
ever, that most of the small granules involved any concretionary 
action after their deposition. They have been considered in the pre- 
ceding paragraphs as fragment al and precipitated granules. The 
peculiarity of the nodules here illustrated is that, though rounded and 
elliptical in outline, the internal texture is granular like the matrix and 
finer parts of the conglomerate. It is evidently a fragmental accumu- 
lation which has been incorporated or included in a nodular cement. 
The nodules are mostly elongated and show gradations into con- 
tinuous beds of similar material, as shown in the figure (Plate VI B). 
Both the rounded forms and the irregular layers contain septaria cracks 
(Plate VIII). Between the nodules are thin beds of magnetite, and the 
nodules themselves contain considerable magnetite, so that the forma- 
tion contains about 20 per cent of iron. 

The conspicuous irregularity in the bedding is probably more closely 
related to these several primary modifications than it is to the original 
precipitation. Solution surfaces resulting from leaching normally show 
rounded forms; concretions are irregularly rounded, and the compres- 
sion of loose sediments against the rounded lumps would no doubt 
produce some such wavy bedding planes. Considerable irregularity 
may be due to the shrinkage of voluminous colloidal precipitates during 
consolidation and recrystallization. The septaria cracks show that 
some such shrinkage occurred. 

Metamorpkism. — It can hardly be assumed that the primary depo- 
sition and reworking of the iron resulted in the formation of magne- 
tite. The precipitation of magnetite in water at ordinary temperature is 
not a reaction commonly observed in nature. There is reason to believe 
that the precipitate was a chert with ferrous silicate and ferrous carbon- 
ate and more or less limonite. The oxidation and formation of intra- 
formational conglomerate probably produced limonite. The change 
from these minerals to magnetite is a kind of metamorphism that may 
occur in several ways. Probably the ferrous iron of the silicate and 
carbonate and the more oxidized limonitic beds reacted directly to 
form magnetite. If the oxidation was very complete, a reducing agent 
is available in the organic matter, known to have been present in cer- 
tain beds. These and other reactions have been sufficient, during the 
long ages of recrystallization under heat and pressure, to make most 
of the iron oxide of the Mesabi formations magnetite. 9 

The occurrence of magnetite and hematite as adjacent pebbles in 
a conglomerate, or as an intergrowth, has already been mentioned, but 
the alteration just outlined offers no explanation of the different state 

8 This is not recognized by Van Hise and Leith. It is not necessary to appeal to the reduction or 
oxidation by water as suggested by them, op. cit., pp. 172 and 527. 



of oxidation of such associated minerals. The ease of alteration of 
ferric oxides to magnetite may be related to certain conditions in the 
primary ore. There might be differences in hydration,- or different 
admixtures of hydrous minerals. There might be differences in poros- 
ity; different amounts of the organic reducing agent; or other pecu- 
liarities of mineral association. Any of these conditions may affect the 
later metamorphism to hematite or magnetite. 

The development of several silicates and the coarse grain of much 
of the taconite may also be attributed to general metamorphism, but 
have had no very great influence on the value of the deposits. 

Dynamic metamorphism is indicated by the folded layers and ten- 
sion cracks. Scattered veinlets of quartz and magnetite probably 
developed under conditions of great heat and pressure. 

The contact metamorphism of the iron formation as a whole has 
been discussed above. It has recrystallized the magnetite near the 
gabbro, increasing the size of grain and ease of magnetic separation. 
It apparently added no appreciable amounts of either iron or titanium. 

Lack of secondary enrichment. — It is likely from the broader rela- 
tions of the gabbro that much of the Virginia slate and even the iron 
formation had been eroded before the gabbro was intruded. 19 There 
may have been enriched portions of iron formation, but if so they have 
probably been eroded since Keweenawan time. The uniformity of the 
iron content in the fresh taconite from east to west on the Mesabi indi- 
cates that before the metamorphism just described there was no enrich- 
ment of the rocks now found on the east Mesabi of any such local sor: 
as that now in progress. Since Keweenawan time enrichment has 
occurred in a few favorable areas. 

The three mines of hematite in the east Mesabi are all in the western 
end of the area, and represent the gradation to the conditions of enrich- 
ment on the main range. East of the Spring mine there is practically 
no effect of weathering. Swamp waters may attack the formation to a 
depth of a few inches, but they impoverish it rather than enrich it. 
The porosity is said to be less than 1 per cent, as compared to 5 per 
cent farther west. 11 This difference in porosity is thought to be more 
important than the mineral composition in determining enrichment, 
for the magnetic oxides are found all across the range, and the mag- 
netite east of Mesaba has in some cases been altered to the ferric form. 

Resume of the history. — Deposition occurred probably in shallow 
water by precipitation, mainly as an organic process, resulting in lean 
ferruginous cherts, with more or less siderite, ferric oxide, and greena- 

» Grout, Frank F., The lopolith: Amer. Jour. Science, vol. 46, pp. 518-20, 1918. 
11 Van Hise and Leith, op. cit., p. 554. 


Contour interval 20 feet 


Duluth gabbro 

Diabase sills and dikes 

Virginia slate 

1 Normal slate 

2 Hornfels(cont3ct equivalent) 

6a Carbonate beds 
(.A Lean thin-bedded taconite, 
cherty and calcareous 

Thin -bedded taconite 
>5%-207° iron in magnetite 
SA Thin-bedded taconite with 
septaria and drag folds 
\&%-2ZJo iron in magnetite 

ft \ 
Aub 4 



3BMassive cherty amphibohte, 

little magnetite 
3ALike 3B but with conspicuous 

fayalite spots 

Intermediate i 

18 Horizon showing I 
'A Magnetite-rich he 

PoUegama quarlzite 

Giants Range granite 

Schist and slate 

Strike and dip 
Horizontal bed 



lite. Alternating with periods of precipitation came periods of solu- 
tion, leaching, oxidation, and wave action, producing intraformational 
conglomerates, granular rocks much richer in iron, and probably some 
layers of pure ferric oxide. 

The richer deposits of magnetite are so characteristically in the 
granule and conglomerate zones that we are led to believe that the 
primary leaching was a determining factor in the development of richer 
magnetites from lean ferruginous cherts, greenalite rock, etc. Later 
when covered with other layers, there may have been more or less 
concretionary rearrangement. Deep burial under slates developed heat 
and pressure that recrystallized a great deal of the formation. The 
iron minerals reacted at this time with each other and with organic 
matter, and possibly with other reducing or oxidizing agents producing 
magnetic oxides of iron. The recrystallization produced shrinkage 
cracks, and some regional movements developed folds, but seem to 
have had no important effects on the formation. Contact action by 
the Duluth gabbro and sills made a great difference in the texture of 
the deposits at the east end. Erosion has exposed the metamorphosed 
beds without any important weathering effects on the minerals. 



The Mesaba area (Plate XII). — The two pits of the Graham mine 
at Mesaba are the most important recent operations of the east Mesabi. 
No detailed examination was attempted, but the walls of both pits 
show septarian concretions indicating that the enrichment occurred 
at the upper magnetite horizon. The algal jasper beds seem to have 
been somewhat resistant and escaped enrichment, for considerable 
amounts appear in the dump. 

The Mayas mine, at the lower magnetite zone, was worked for some 
hematite of fair but variable quality. It was reopened in 1918 for some 
manganese that was recently discovered. 

One of the best exposures of algal structure appears in the jasper 
northeast of the town along the road. 

Greenalite, the amorphous green ferrous silicate considered by Leith 
to be the primary precipitate of most of the iron formation, was found 
in its best development in the material on the dump of a deep well at 
the north side of town. Many test pits at the south of the Mayas 
mine have dumps of rock resembling greenalite. 

South of the area mapped in detail, in sec. 34, T. 59 N., R. 14 W., is 
a drill hole, reported by Van Hise and Leith 1 to penetrate some slate 
and then, passing 576 feet of iron formation, to go a number of feet 
into diabase. Diabase outcrops were not seen west of a thin sill 
exposed on the south side of sec. 15, T. 59 N., R. 14 W. 

Magnetite rock that appears to have over 20 per cent of iron in 
magnetic form outcrops in both the upper and lower horizons. The 
lower horizons are well exposed in the N.E. 34 sec. 15, T. 59 N., R. 14 W. 
The higher horizon is exposed near the road, along the south side of sec. 
15 and the north half of sec. 22, T. 59 N., R. 14 W., and seems to represent 
the same body of magnetite as the larger exposures farther east. 

No detailed work has been done south of the iron formation, but the 
approximate contact of the slate and gabbro south of Mesaba is easily 
determined by the outcrops along the railroad. 

The Spring mine area (Plate XIII). — The Spring mine lies in a 
valley between hills of Lower-Middle Huronian slates on the west, 
and a bluff of slate and slaty taconite on the east. The ore has an 
easterly dip, and was enriched down even under the slate outcrops. 
The floor of the ore body is partly the Pokegama and partly older for- 
mations. The concentration and enrichment are discussed in Chapter V. 

i Van Hise, C. R., and Leith, C. K., The geology of the Lake Superior region: U. S. Geol. Survey 
Mon. 52, p. 177. 





The old pit was pumped out in 1917 for further exploration. It is 
referred to as the "Silverton property," but is locally still called the 
Spring mine because of a strong spring that issues near by. 

The casing of an old drill hole in the western part of sec. 14, 
T. 59 N., R. 14 W., has a flow of water from artesian pressure. The beds 
no doubt outcrop in the hill half a mile north. The water is very clear 
and apparently not ferruginous. 

The fold in this area is the largest in the east Mesabi, and may 
have been a factor in determining the location of the ore of the Spring 
mine. It is close to the change in bedrock below from granite to. slate, 
but the slate stands at a higher level than the granite under the iron 
formation. There are intrusives which may be related to the struc- 
ture. The outcrops may be interpreted as the fold shown on the map 
and block diagram. The steep dips of the beds in the narrow part of 
the fold, along the railroad, show that the folding was very sharp. It 
is possible that some faulting may have occurred also. Small faults 
may be seen in the slate. The discovery of some outcrops far to the 
south along the east side of sec. 23 makes the curve considerably 
sharper than it was previously supposed to be. (See Figure 6.) 

It is in the northern part of this area that the Giants Range granite 
has been supposed to give way to the Embarras granite, a later intru- 
sive on the east. No such intrusive could be found. 

The exposures along the railroad south of Spring mine include sev- 
eral that were the bases of interpretation of the geologic history. In 
the N.E. Y± of N.E. H sec. 14, T. 59 N., R. 14 W., a few feet east of the 
railroad, iron formation conglomerate lies on the eroded edges of Huronian 
slate, with a striking unconformity. 

In the S.W. 34 oi S.W. of "the same section, the railroad passes 
some outcrops of slaty beds that resemble a phase of Virginia slate. 
The outcrops, however, can be followed directly west in a belt that 
passes between beds of iron formation, and they are undoubtedly outcrops 
of the intermediate slate. The slaty outcrops have conspicuous garnet 
metacrysts, which under the microscope prove to be filled with minute 
amphibole needles. 

The magnetite bodies show some features that differ from those 
farther east. The upper magnetite horizon is relatively narrow, but 
the lower, by reason of increased thickness and relatively flat variable 
dip, shows a broad important area. A preliminary sampling of the 
area indicates over 27 per cent of iron in form for magnetic concen- 
tration, in an area of almost half a square mile. The thickness of this 
lower formation is indicated by some drill holes in the next sections 
east of Spring mine. They show over 40 feet of this lower bed, con- 
taining about 30 per cent of iron. Drilling was done in both sec. 1 


Contour interval 20 feet 


Duluth gabbro 

Diabase sills and dikes 

Virginia slate 
Normal slate 
I Hornfelsfcontact equivalent) 

6a Carbonate beds 
oA Lean thin-bedded taconil 
cherty and calcareous 

Aub .. 

5d Thin-bedded taconite 
l5%-20% iron in magnetite 

5A Thin-bedded taconite with 
septa ria and drag folds 
m%-22y<, iron in magnetite 

seThin horizon with bowl£lgaJ?) 

3BMassive cherty amphibolite, 

little magnetite 
!ALike3B but with conspicuous 

fayalite spots 

Intermediate slate 


IC Massive cherty taconite, 

little magnetite 
16 Horizon showing bowl (algal ?) 

IA Magnetite-rich horizon 

Pokegama quajuzite 

Giants Range granite 

Schist and slate 

Strike and dip 

Horizontal bed 



md sec. 12, T. 59 N., R. 14 W., and indicates 50 feet or more of the 
'tipper body of over 20 per cent magnetite, and from 10 to 61 feet of 
the lower body in different places. Most of the lower magnetite body 
lies close to the granite. Plate XIV A, showing the structure, is based 
on accurate measurements of drilling. 

The Ridge area {Plate XV). — The branch of the railroad running 
along the east Mesabi formerly had a side line turning north just east 
of Ridge. This passed through a valley transverse to the range. On 
each side of this low place the exposures of the several divisions of 
iron formation are very clear. Through most of the four sections 
mapped, the outcrops are numerous and horizons easily distinguished. 
Along the south side, however, the limit of the formation is concealed 
in swampy land. 

A few paces east of the center of sec. 32, T. 60 N., R. 13 W., is a small 
ravine running south through the richest part of the magnetite beds and 
the intermediate jasper with algal structure. A sample from a bluff 10 
feet high showed nearly 40 per cent of iron in form for magnetic con- 

The northern, lower magnetite belt here seems to become less con- 
tinuous than it is west of Spring mine, though the several outcrops may 
be connected under the drift. The lower magnetite beds show more 
oxidation and leaching than in most outcrops. The lower slaty beds 
have many soluble silicates and are similarly leached and enriched, 
developing oolitic or concretionary structures with limonite composition. 

A small diabase dike intruding the lower horizons of the Biwabik 
formation, outcrops in the N.W. }/i of the N.E. 34 of sec. 32. 

Over 60 drill holes were once put down in sec. 6 and adjoining parts 
of sec. 5 and sec. 7., T. 59 N., R. 13 W., and the cores have been recently 
- eviewed. They can be correlated closely with the data from outcrops 
and other drilling. The upper magnetite body is 212 feet thick and 
^he lower about 42 feet thick. (See Figure 4 and Plate XIV.) 

The Jericho area. — Two properties in this area have been more or 
less prospected. Many test pits were put down a little northwest of 
the center of sec. 28, T. 60 N., R. 13 W., northwest of the old "Syndi- 
cate camps. " The dumps show a mixture of magnetite and ferric 
oxides, but enrichment has not apparently yielded a large rich body of 
ore. The compact specimens on the dump indicate that leaching has 
not been as thorough as in many places farther west, though evidently 
silica has been leached. Within a mile southeast of these prospects, 
in a swampy area, there is a highly ferruginous spring, indicating that 
iron as well as silica is being removed from the rocks. 

The other prospect is the recent opening in magnetite in a bluff 
south of the railroad siding at Jericho. This bluff is in the gradation 


zone between good magnetite deposits north and the lean taconite 
south. The tests were made on quarry samples and were not good 
enough to encourage further work. The exposure can be followed 
southwest nearly a mile along the strike, with fairly uniform char- 
acter. North of the railroad are a few outcrops of more promising 

Through most of this area the outcrops are few, and the drift is 
apparently thick. In all of sec. 27, T. 60 N., R. 13 W., only two 
small outcrops were found. These are of the lean upper beds. This of 
course makes the boundaries as mapped somewhat uncertain. 

The Iron Lake area (Plate XVI). — The southern part of this area 
is characterized by a series of diabase ridges formed by outcropping 
sills 10 to 20 feet thick. These can be followed with considerable accu- 
racy for a mile or more, and extend with some interruptions much 
farther. The sills conform to the bedding wherever observed and can 
be used in some cases as a guide to the structure of the other beds. 
The dip of the sills is so slight in some places that the topographic 
slope exceeds the dip, and there are numerous outliers and inliers out- 
cropping in a way which interferes somewhat with the regularity of the 
belts mapped. The sills are the only conspicuous outcrops in the south- 
ern area, and it is only in a few places that their roof and floor rocks 
are well exposed. The whole series has been previously interpreted as 
a single sill, half a mile wide, but there are at least four in this area. 
The tops of several of these diabase ridges are a hornfels, clearly derived 
from slate. Without much question the Virginia slate is continuous 
from Mesaba to Dunka River. 


SIDE OF N.E. yi SEC. 25, T. 60 N., R. 13 W., SHOWING OUTCROPS AND (IN 

Two parts of this area show sharp and exceptional folds a few hun- 
dred feet wide. One is in the southwest corner of sec. 26, and one in 
the N.E. y± sec. 25, T. 60 N., R\ 13 W. It is assumed from the out- 
crops in these places that the folds involve the beds above and below 
to some extent. Figure 7 shows the outcrops in sec. 25, in a cross sec- 
tion interpreting them as two folds. Such a structure, however, may 
have resulted from displacement along two faults. Figure 8 shows 
the similar structure in sec. 26. 


Dxduth gabbro 

Diabase sills and dikes 

H ornfelsfcontact equivalent; 

Carbonate beds 
Lean thin -bedded taconite, 
cherty and calcareous 

Thin-bedded taconite 
i5%-20% iron in magnetite 
Thin-bedded taconite with 
septaria and drag folds 
\Q%-ZZJ a iron in magnetite 

Thin horizon with bowlegs)?) 

Conglomeratic beds 
18^.-35% iron in magnetite 

Massive cherty amphibolite, 
Jt * little magnetite 

Like 38 but with conspicuous 
A fayalite spots 

Intermediate slate 

Poke gam a quarl 

Giants Range granite 

Contour interval 20 feet 

Schist and slate 

Strike and dip 

Horizontal bed 



Contour interval 20 feet 


Duluth gabbro 

Diabase sills and dikes 

Virginia slate 
j. Normal slate 
2 Hornfels(contact equivalent) 

Carbonate beds 
bp, Lean thin-bedded taconite, 
cherty and calcareous 

Aub s 

5Q Thin-bedded taconite 
P5%-207« iron in magnetite 

5A Thin-bedd3d taconite with 
septaria and drag folds 
I8^-2Z7» iron in magnetite 

4gThin horizon with bowl^lgalr) 

4,\ Conglomeratic beds 

l8/»-35% iron in magnetite 

3S Massive cherty amphibolite, 
little magnetite 
Like 3B but with conspicuous 
fayalite spots 

Intermediate slate 


IC Massive cherty taconite, 

little magnetite 
IB Horizon showing bow I (algal .9 

1A Magnetite-rich horizon 

Poke gam a quart zite 

Giants Range granite 

Schist and slate 

Strike and dip 

Horizontal bed 



In the Sulphur area there has been considerable exploration of a 
magnetite body which has been followed into this area in sec. 24, T. 
60 N., R. 13 W., with little apparent change in character. In sec. 26 
it is probably similar, but outcrops are not so numerous. 


A strong magnetic pole appears in the magnetic map of this area, 
located in the east central part of sec. 24. The dip needle gives fairly 
high readings and the compass needle points in from all sides with 
surprising regularity. These results were not duplicated anywhere 
along the range and were thought worthy of a drill exploration. Four 
holes were put down about 50 feet in rock. Certain layers proved to 
be rich, but one of the holes struck a concealed mass of diabase, and 
as a whole the drilling failed to explain the magnetic attraction. The 
explanation may possibly be found in the topographic situation, — 
there is a deep valley cutting the beds, — or in the folded or faulted 
structure, described above. 

The Sulphur area (Plate XVII). — The outcrops in sees. 18 and 19, 
T. 60 N., R. 12 W., have long been noted as rich in magnetite. They 
are located on a hill slope that spreads them out to a greater width 
than in most of the areas. The irregularity in the belt is due more to 
the topography than to folding. Practically the whole south slope is 
a conspicuous outcrop of conglomerate taconite. The amount of iron 
is about the same as that in the equivalent beds all along the range. 
Very little hematite is to be found in this area. This is attributed, 
probably correctly, to the proximity of the gabbro on the southeast. 

Along the boundary of the iron formation and the granite, the con- 
tact shows more clearly than elsewhere the unconformity between the 
Upper Huronian and older formations. Not only is there a prominent 
conglomerate in many places, but there are exposures enough to out- 
line an older erosion surface of considerable relief. Figure 2, page 6, 
is a section along the general trend of the contact, across the line 
between sees. 17 and 18, T. 60 N., R. 12 W. 



The diabase belt on the map is traced by outcrops well across into 
sec. 20, but a second sill is revealed by the drilling. The sill between 
Aub 5 and Aub 6 has been penetrated in several places by recent drill- 
ing. The characteristic texture of Aub 5 appears in the core just below 
the diabase, but there are 20 to 40 feet of this in which the magnetic 
iron content is less than 20 per cent. It is evident that the upper part 
of Aub 5 is leaner than the bottom, though noticeably of better grade 
than beds Aub 6 . 

South of the diabase in sec. 20 there are outcrops of gabbro and 
altered slate. This is one of the best places along the contact of the 
gabbro and slate to observe the contact effects of the gabbro. The 
slate develops the hornfels texture and the appearance of a fine-grained 
gabbro. There are phases that consist of fragments in a matrix, as if 
part of the slate had been almost a liquid and gathered in some inclu- 
sions. Pegmatitic stringers and many contact minerals appear in the 

Drilling has been carried over the magnetite body almost all across 
sees. 17 and 19, T. 60 N., R. 12 W. The exploration has here demon- 
strated a fair degree of uniformity of the beds, as suggested by their 
outcrops, but a lean area is found at the east. About 150 holes have 
been put down, most of them to a depth of 50 feet in rock, but others 
to greater depths, to determine the general relations. The drilling 
operations by two or three companies at different periods have been 
reviewed and correlated. Several holes were placed for the sake of 
determining the geological horizons and thickness of the formations. 
No single hole, however, has been carried through the whole forma- 
tion. The section shown in Plate XIV B is based mostly on outcrops, 
topography, and surface measurements of the inclination of the beds. 

Two quarries have been opened in the magnetite bodies, and many 
cars of ore have been shipped to Duluth for concentration. 

The Birch Lake area {Plate XVIII). — Between the area at Sulphur 
and Dunka River there are few outcrops. Those that occur along the 
granite boundary indicate the continuity of the belt of iron formation. 
On the southeast, there are outcrops of gabbro, turning north, some- 
what abruptly, and at the river there is gabbro within a quarter of a 
mile of the granite. From Dunka River to Birch Lake outcrops are 
numerous and mapping can be done in considerable detail. Gabbro 
and granite and intervening sediments are all clearly exposed. 

The contact effects of the gabbro extend all through the sediments 
in this area and obliterate most of the distinctions between beds. It is 
here that the bands of magnetite and quartz show the strong contrast 
in color, suggesting some segregation and intensification of banding 
during contact action. There is evidence of considerable rearrange- 



ment, in a thin section of pyroxenite from this area; a black oxide 
clearly replaces parts of the pyroxene. 

Certain beds have been recognized, however. Distinguished from 
the leaner beds, the beds with over 30 per cent iron in magnetite can 
be recognized in the central zone of the formation almost to the end 
of the outcrops of iron formation, in sec. 26, T. 61 N., R. 12 W. The 
conglomerate horizon could be distinguished nearly everywhere that 
iron formation was seen, in sjpite of the fact that it was well impreg- 
nated with pegmatitic stringers from the gabbro. There are also in 
this area beds which can be correlated with the lower cherty algal beds, 
the intermediate slate, and the upper lean beds just below the Virginia 
slate. The sediments, from Dunka River north, are steeply and not 
very regularly tilted, and the strike is variable. Figure 9 is a cross 
section in sec. 3, T. 60 N., R. 12 W., indicating that the iron forma- 
tion has been reduced in thickness to 360 feet. The exposed bodies of 
magnetite are very narrow. 


Where the exposures are clear the four main divisions of the iron 
formation can be recognized and can be correlated with those in the 
areas farther west. Above the lean upper beds of iron formation are 
some fine grained rocks with a texture and mineral character identical 
with the hornfels developed from Virginia slate, south of Sulphur Sid- 
ing. The slate is therefore mapped here much farther north than it 
has heretofore been shown. It has no slaty character, but is easily 
distinguished from the coarser gabbro in most places. The gabbro 
apparently cut down a little irregularly into the sediments, for it 
reaches through the slate to the iron formation at one place in sec. 3, 
and another in sec. 35. The last large outcrop of hornfels (slate) is 
farther northeast in sec. 35. 



Several diabase sills and dikes intrude the sediments near the upper 
beds of iron formation. The sills are porphyritic, but where the enclos- 
ing sediments are altered to hornfels, the sills also become hornfels. 
They contain phenocrysts which are so characteristic that one may 
identify them with confidence as the equivalent of the diabase sills in 
the Iron Lake area; but the sills lose practically all trace of their orig- 
inal diabasic texture. 

In the northeast corner of sec. 26 outcrops of gabbro and granite 
come together, definitely pinching out all sediments. The iron forma- 
tion was formerly mapped as extending to Birch Lake, probably 
because outcrops are not abundant, and the iron formation can be 
found again near the granite north of Birch Lake. This patch north 
of the lake must be considered an isolated remnant like those farther 
east along the gabbro contact, 2 for the continuous outcrops of iron 
formation definitely end south of the lake. 

Broderick, T. M. f Titaniferous magnetites of Minnesota: Econ. Geology, vol. 12, pp. 663-96, 1917. 



Algal deposits 17, 21, 22, 36, 40, 48, 51 

Algonkian system 5 

Alternation of deposition 33, 34, 42 

Amphibole-magnetic rock 15, 20, 30 

Amphiboles 9, 11, 12, 29, 30, 37, 49 

Animikie group 7-8 

Apatite 14 

Arsenopyrite 14 

Augite 13 

Aurora, ore deposits near 32 

Babingtonite 13 

Bedding planes 22-23, 37, 44 

Birch Lake 1, 30, 32, 36, 54, 56 

Biwabik iron formation 5, 7, 17, 34, 41 

correlation chart 24 

tabular section 14 

Breccia 17 

Broderick, T. M., referred to 2, 34, 56 

Burchard, E. F., referred to 35 

Carbonates 12, 25 

Chalcopyrite 14 

Chemical composition of magnetite rock 29 

Chert 3, 7, 12, 14, 17, 18, 20, 26, 30 

Chester, A. H., referred to 2 

Chlorite 14, 17 

Clarke, F. W., referred to 18 

Coleraine 10, 32 

Commercial situation 35 

Conchoidal fracture 19 

Concretions 30, 44, 45, 47, 48 

Conglomerate 16, 20, 22, 44 

Conservation 39 

Contact metamorphism 8-10, 47, 54 

Correlation of east and west Mesabi . 26 

Cronk, referred to 3 

Cross sections 

Spring mine 50 

Iron Lake area 52, 63 

Birch Lake area 55 

Cuyuna iron formation 41 

Davis, E. W., referred to 23 

Deposition of iron 40-41 

Diabase 8, 3 1, 48, 52, 53, 54, 56 

Dikes 8, 56 

Diopside 13 

Dip 34,35 

Distribution of magnetite 32 

Districts, detail of 47 

Drift 32,36 

Duluth and Iron Range Railroad 38 

Duluth gabbro. See Gabbro 8 

Dumps, location 39 

Dunka River 8, 10, 23, 30, 38, 52, 54 

Eames, Henry H., referred to 2 

Economic considerations 35-39 

Embarras granite, absence of 49 

Enrichment of hematite ores 27 

lack of muck 46 

Enrichment of ores 27, 43, 46 


Epidote 14 

Estimating reserves 34 

Eveleth, iron deposits near 32 

Experiment Station, Minnesota School 

of Mines 29 

Exploration, cost 39 

Exploration of east Mesabi 3 

Faults 49,52,53 

Fayalite 11, 12, 19, 32, 37 

Feldspars 13 

Ferric oxide 3 

Foerste, A. G., referred to 43 

Formations 5 

Gabbro 5, 7, 8, 9, 13, 29, 47, 53, 54, 55 

Garnet 13, 17, 23 

Geology of East Mesabi Range 5-11 

Giants Range 5, 37 

Gilbert, iron deposits near 32 

Glaciation 1 1 

Graham mine 27, 48 

Grantie, Giants Range 5 

Granule textures 29 

Graphite 13 

Graphitic cherts 40 

Greenalite 13, 14, 30, 40, 48 

Grout, Frank F., referred to 2 

Gunflint Lake 29 

Harder, E. C, referred to 41 

Hardness of taconite 38 

Hematite 12, 14, 26, 27-28, 45, 46, 48, 53 

Hematite ore deposits 27 

Hibbing, iron deposits near 32 

History of deposition 46 

History of the region 1 

Hornfels 9,52,54,56 

Hudson's Bay, iron formations near 18 

Huronian series 5, 48 

Hypersthene 13 

Intraformational conglomerate 42, 45 

Introduction l 

Intrusions 31, 49 

Iron Lake 34,52-53 

Iron oxide 29 

Iron, source of 40 

Jasper 12,21,40,48 

Jericho area 51-52 

Johnston, A. W., referred to 41 

Jordan, Fred, referred to 31 


Keweenawan series 5, 8, 46 

Kingston, referred to 3 

Leaching 12,26, 43,44, 45 

Leith, C. K., referred to 2, 3, 5, 8, 19, 

25, 26, 29, 30, 33, 35, 40, 41, 

43, 48 

Lenticular beds 22, 23 

Limestone 25 

Limonite 11, 12 

Location of district 1 



3 5002 03596 5230 



Lower cherty beds 16 

Lower slaty beds 16 

Lower-middle Huronian series 6 

McCarthy. J. H., analyses by 29 

Magnetic iron 31 

Magnetic mapping 33-34, 53 

Magnetic tests 39, 53 

Magnetism 37 

Magnetite deposits 29-34 

distribution of 32 

economic consideration 35-39 

origin 3-4, 40-47 

Malachite 14 

Manganese ore at Mayas mine 48 

Mayas mine 27, 48 

Mesaba 1, 52 

Mesabi Syndicate 2, 23, 38 

Metacrysts 15 

Metamorphism 45-46, 47 

contact 8-10,47, 54 

regional 10 

Mica 13 

Milling 38 

Minerals of the iron-bearing formation 12 

Minerals of the magnetite deposits 27 

Mining methods 35-36 

Minnesota School of Mines, Experiment 

Station 29 

Modification of deposit, primary 41 

Moore, E. S., referred to 18 

Muscovadite 9 

Oliver Iron Mining Company 3 

Open pit mines 27 

Ore, grade 35-37 

Origin of magnetite 3-4, 40-47 

Pegmatite stringers 9, 12, 54, 55 

Per cent of iron 35-37, 48 

Phosphorus 29, 38 

Pleistocene deposits 5 

Pokegama quartzite 7, 18, 25, 34, 48 

Precipitation 40, 41 

Production of iron ore 35 

Pyrite 14 

Pyroxenes 13 

Pyroxenite 9. 15, 55 

Pyrrhotite 14 

Quality of product, (grade) 38 

Quartz 11, 12 

Quartzite 15 

Quaternary system 5 


Regional metamorphism 10 

Reserves, method of estimating 34 

Ridge area 51-52 

Rocks of the iron-bearing formation 13 

Sampling 39 

Scale of operations 38 

Sebenius, J. U., referred to 3 

Sedimentation, rhythmic 42 

Septaria 23,30,45,48 

Siderite 32 

Sills 8,52,54,56 

"Silverton property" 49 

Sinter 38 

Size of grain 38 

Slate 5. 7, 9, 17,52,54, 55 

Slaty beds of iron-bearing formation 17, 19 

Source of iron of deposits 41 

Spring mine 23, 27, 46, 48-51 

Spurr, J. E., referred to 2 

Steiger, George, referred to 19, 40 

Stratigraphy of magnetite 31-32 

Structure of iron formations 10, 37 

Sub-divisions of iron-bearing formation 14 

Sulphur area 30, 31, 53-54 

"Syndicate camps" 51 

Taconite, defined 13 

Taconite, described 14-26 

composition of 29 

textures of 30, 41-42, 44 

Tailings, soluble iron in 32 

Textures of magnetite 29-31, 41-42 

Titanium 29 

Topography 37 

Tourmaline 14 

Transportation of ore 38 

Upper cherty beds 21 

Upper slaty beds 24 

Upper Huronian series 6 

Van Hise, C. R., referred to 2,*5, 11, 19, 


Virginia, iron deposits near 32 

Virginia slate 7, 17, 52, 55 

Volume of ore 36 

Wash ores 36 

Walcott, Charles D., (quoted) 21-22 

Weathering 11 

Winchell, H. V., referred to 2 

Winchell, N. H., referred to 2, 8, 9 

Wisconsin age 5 

Wolff, referred to 3, 16, 17, 18, 25, 26, 27, 32 


ss 3 5002 03596 5230 


Lower cherty beds 16 

Lower slaty beds 16 

Lower-middle Huronian series 6 

McCarthy. J. H., analyses by 29 

Magnetic iron 31 

Magnetic mapping 33-34, 53 

Magnetic tests 39, 53 

Magnetism 37 

Magnetite deposits 29-34 

distribution of 32 

economic consideration 35-39 

origin 3-4, 40-47 

Malachite 14 

Manganese ore at Mayas mine 48 

Mayas mine 27,48 

Mesaba 1, 52 

Mesabi Syndicate 2, 23, 38 

Metacrysts 15 

Metamorphism 45-46, 47 

contact 8-10, 47, 54 

regional 10 

Mica 13 

Milling 38 

Minerals of the iron-bearing formation 12 

Minerals of the magnetite deposits 27 

Mining methods 35-36 

Minnesota School of Mines, Experiment 

Station 29 

Modification of deposit, primary 41 

Moore, E. S., referred to 18 

Muscovadite , 9 

Oliver Iron Mining Company 3 

Open pit mines 27 

Ore, grade 35-37 

Origin of magnetite 3-4, 40-47 

Pegmatite stringers 9, 12, 54, 55 

Per cent of iron 35-37, 48 

Phosphorus 29, 38 

Pleistocene deposits 5 

Pokegama quartzite 7, 18, 25, 34, 48 

Precipitation 40, 41 

Production of iron ore 35 

Pyrite 14 

Pyroxenes 13 

Pyroxenite 9, 15, 55 

Pyrrhotite 14 

Quality of product, (grade) 38 

Quartz 11, 12 

Quartzite 15 

Quaternary system 5 


Regional metamorphism 10 

Reserves, method of estimating 34 

Ridge area 51-52 

Rocks of the iron-bearing formation 13 

Sampling 39 

Scale of operations 38 

Sebenius, J. U., referred to 3 

Sedimentation, rhythmic 42 

Septaria 23, 30, 45, 48 

Siderite 32 

Sills 8,52,54,56 

"Silverton property" 49 

Sinter 38 

Size of grain 38 

Slate 5,7,9, 17,52,54,55 

Slaty beds of iron-bearing formation 17, 19 

Source of iron of deposits 41 

Spring mine 23, 27, 46, 48-51 

Spurr, J. E., referred to 2 

Steiger, George, referred to 19, 40 

Stratigraphy of magnetite 31-32 

Structure of iron formations 10, 37 

Sub-divisions of iron-bearing formation 14 

Sulphur area 30, 31, 53-54 

"Syndicate camps" 51 

Taconite, defined 13 

Taconite, described 14-26 

composition of 29 

textures of 30. 41-42. 44 

Tailings, soluble iron in 32 

Textures of magnetite 29-31, 41-42 

Titanium 29 

Topography 37 

Tourmaline 14 

Transportation of ore 38 

Upper cherty beds 21 

Upper slaty beds 24 

Upper Huronian series 6 

Van Hise, C. R., referred to 2, 5, 11, 19, 


Virginia, iron deposits near 32 

Virginia slate 7, 17, 52, 55 

Volume of ore 36 

Wash ores 36 

Walcott, Charles D., (quoted) 21-22 

Weathering 11 

Winchell, H. V., referred to 2 

Winchell, N. H., referred to 2, 8, 9 

Wisconsin age 5 

Wolff, referred to 3, 16, 17, 18, 25, 26, 27, 32